# Booking.com Data Science

The Machine learning and Data Science blog of Booking.com - Medium

This is one page of public article previews, not the complete archive. Follow Next page to continue. Summaries are not the original full articles.

## From Static Help Content to a RAG-Based Onboarding Agent for Partners

DevFeed: [From Static Help Content to a RAG-Based Onboarding Agent for Partners](<https://devfeed.tech/articles/from-static-help-content-to-a-rag-based-onboarding-agent-for-partners-89968.md>)

Original publisher: [Read original article](<https://booking.ai/from-static-help-content-to-a-rag-based-onboarding-agent-for-partners-f7d5a789c4f7?source=rss----4d265f07defc---4>)

Author: Anastasiia Barchukova

Published: 2026-10-19T09:35:24Z

Content type: article

Language: en

Sources: [Booking.com Data Science](<https://devfeed.tech/sources/booking-com-data-science.md>)

Topics: [Retrieval Augmented Generation (RAG)](<https://devfeed.tech/topics/retrieval-augmented-generation-rag.md>), [AI Chat](<https://devfeed.tech/topics/ai-chat.md>), [Conversational AI](<https://devfeed.tech/topics/conversational-ai.md>), [AI-driven information retrieval](<https://devfeed.tech/topics/ai-driven-information-retrieval.md>), [AI search](<https://devfeed.tech/topics/ai-search.md>), [managing large-scale text data in AI](<https://devfeed.tech/topics/managing-large-scale-text-data-in-ai.md>), [human review](<https://devfeed.tech/topics/human-review.md>)

Tags: [account](<https://devfeed.tech/tags/account.md>), [agent](<https://devfeed.tech/tags/agent.md>), [ai](<https://devfeed.tech/tags/ai.md>), [ai-agent](<https://devfeed.tech/tags/ai-agent.md>), [architecture](<https://devfeed.tech/tags/architecture.md>), [backend](<https://devfeed.tech/tags/backend.md>), [booking](<https://devfeed.tech/tags/booking.md>), [business](<https://devfeed.tech/tags/business.md>), [chatbot](<https://devfeed.tech/tags/chatbot.md>), [content](<https://devfeed.tech/tags/content.md>), [context](<https://devfeed.tech/tags/context.md>), [conversational-ai](<https://devfeed.tech/tags/conversational-ai.md>), [featured](<https://devfeed.tech/tags/featured.md>), [help](<https://devfeed.tech/tags/help.md>), [onboarding](<https://devfeed.tech/tags/onboarding.md>), [partner](<https://devfeed.tech/tags/partner.md>), [partners](<https://devfeed.tech/tags/partners.md>), [platform](<https://devfeed.tech/tags/platform.md>), [process](<https://devfeed.tech/tags/process.md>), [quality](<https://devfeed.tech/tags/quality.md>), [questions](<https://devfeed.tech/tags/questions.md>), [rag](<https://devfeed.tech/tags/rag.md>), [responses](<https://devfeed.tech/tags/responses.md>)

### AI overview

Booking.com's onboarding agent improved first-line task handling by fixing retrieval and routing issues rather than changing prompts. Audits found navigation noise, overly small chunks, and irrelevant partner-specific material in the help index; the ingestion process was updated to filter content, refine chunking, and expand retrieved context. The agent also uses separate retrievers for registration guidance and legally restricted questions, which receive pre-approved static answers instead of generated responses. In an online experiment scored across roughly 2,000 conversations, task-completion scores rose 1.33x, with answer relevance, refusal rate, and response reliability also improving. The task-completion measure reflects judge-assessed handling within scope, not confirmed end-to-end resolution; about 15% of older chatbot conversations lacked a required registration-status field and were excluded from the primary comparison.

### Source excerpt

Authors: Anastasiia Barchukova, Antonio Castelli, Artem Alimov, Kate Rasskazimova, Niels Ranzijn, Susanna Agababyan Every property on Booking.com is listed by a partner -- the person or business that owns or manages the place to stay. Some partners are self-managed: they sign up, list their property, and run it entirely on their own, without a dedicated account manager. For them, the registration process -- a multi-step flow where they create an account and describe their property -- is often where they get stuck, due to its length, unfamiliar terminology, and business-critical decisions they must make before they've learned how the platform works. To help self-managed partners get through registration on their own, we built a chatbot that answers their questions as they come up, without needing a human account manager. Our chatbot should therefore be deeply grounded in the context of the registration funnel, while also being capable of managing partner expectations by answering broader questions about the platform and what comes after registration. That is the problem this article is about. It describes the approach behind the next iteration of the solution: an agent embedded directly in the registration flow. By design, the assistant is scoped to answering partners' questions in context, grounded strictly in the partner help center content -- it does not generate free-form responses outside that knowledge base. The system at a glance At a high level, the chatbot sits inside the registration flow, routes requests through a dedicated backend layer, retrieves relevant help-center content, and uses the conversational AI platform to generate grounded responses. Figure 1 illustrates this simplified end-to-end architecture. Figure 1: Simplified architecture of the onboarding chatbot, showing the registration flow, chatbot service, help-content retrieval, response generation, and the feedback loop used to improve quality over time.Diagnosing before touching anything Already i

## Introducing AgentHub: Booking.com Platform for Modular Travel AI Agents

DevFeed: [Introducing AgentHub: Booking.com Platform for Modular Travel AI Agents](<https://devfeed.tech/articles/introducing-agenthub-booking-com-platform-for-modular-travel-ai-agents-64207.md>)

Original publisher: [Read original article](<https://booking.ai/introducing-agenthub-booking-com-platform-for-modular-travel-ai-agents-68bc521a0f50?source=rss----4d265f07defc---4>)

Author: Moran Beladev

Published: 2026-10-02T16:12:51Z

Content type: article

Language: en

Sources: [Booking.com Data Science](<https://devfeed.tech/sources/booking-com-data-science.md>)

Topics: [Artificial Intelligence](<https://devfeed.tech/topics/ai.md>), [agentic workflows](<https://devfeed.tech/topics/agentic-workflows.md>), [Large Language Model](<https://devfeed.tech/topics/llm.md>), [AI Chat](<https://devfeed.tech/topics/ai-chat.md>)

Tags: [agent-management](<https://devfeed.tech/tags/agent-management.md>), [agentic](<https://devfeed.tech/tags/agentic.md>), [agentic-ai](<https://devfeed.tech/tags/agentic-ai.md>), [agents](<https://devfeed.tech/tags/agents.md>), [ai](<https://devfeed.tech/tags/ai.md>), [ai-agents](<https://devfeed.tech/tags/ai-agents.md>), [apis](<https://devfeed.tech/tags/apis.md>), [automated](<https://devfeed.tech/tags/automated.md>), [autonomous](<https://devfeed.tech/tags/autonomous.md>), [deploy](<https://devfeed.tech/tags/deploy.md>), [featured](<https://devfeed.tech/tags/featured.md>), [genai](<https://devfeed.tech/tags/genai.md>), [language-models](<https://devfeed.tech/tags/language-models.md>), [large-language-models-llms](<https://devfeed.tech/tags/large-language-models-llms.md>), [llms](<https://devfeed.tech/tags/llms.md>), [machine-learning](<https://devfeed.tech/tags/machine-learning.md>), [modular](<https://devfeed.tech/tags/modular.md>), [monitoring](<https://devfeed.tech/tags/monitoring.md>), [natural-language](<https://devfeed.tech/tags/natural-language.md>), [planner](<https://devfeed.tech/tags/planner.md>), [platform](<https://devfeed.tech/tags/platform.md>), [travel](<https://devfeed.tech/tags/travel.md>)

### AI overview

Booking.com's AgentHub is a configurable platform for building and operating travel agents from reusable templates and shared infrastructure. Its Python SDK configures each agent's model, tools, prompt, execution strategy, and output schema. Platform features include OpenTelemetry traces, continuous LLM-based evaluation, guardrails such as prompt-injection detection and PII filtering, configuration management, and reliability monitoring. In an AI Trip Planner property Q&A flow, the agent resolves property mentions, retrieves current information, accesses structured property data, and uses geospatial tools. The authors report that replacing a deterministic flow with this agentic implementation improved answer relevance by 53%, context relevance by 15%, and factual accuracy by 3.3%, while reducing LLM rejection rate by 64%. They describe multi-agent collaboration, multimodal inputs, long-term memory, and harness optimization as future work.

### Source excerpt

Authors: Moran Beladev, Hadas Harush, Chana Ross At Booking.com we developed AgentHub, a platform to power intelligent, modular, and multi-modal travel experiences through LLM-driven agents. Our platform provides configurable agents capable of performing a wide range of tasks, along with out-of-the-box agent management and monitoring capabilities. With AgentHub, different teams can build and deploy AI agents in a matter of hours, based on configuration and reusable components. Background As AI capabilities continue to advance rapidly, agentic AI is becoming a defining force in the technology landscape. Unlike traditional generative systems, agentic AI extends large language models (LLMs) with iterative planning, autonomous decision-making, and the ability to take action through tools. By enabling direct interaction with internal systems and APIs, these agents can leverage proprietary data and orchestrate workflows end to end -- unlocking deeper contextual understanding and highly personalized user experiences. The Evolution from Generative AI to Agentic AI (source) To support a wide range of GenAI use cases at Booking.com, from simple LLM-powered features to fully agentic capabilities, we built the AgentHub platform. AgentHub enables rapid agent creation through configuration-driven setups, backed by robust shared infrastructure. This approach democratizes GenAI across the company, empowering diverse roles such as UX writers, product managers, developers, and more to build and iterate quickly. The platform supports use cases across the entire booking funnel. From the Discover & Plan stage, where travelers can chat with an AI trip planner to explore and plan itineraries, to Search & Book, enabling natural-language search, smart filtering, and automated property descriptions, and finally Cross-Sell & Trip Management, with agents like the car rental assistant and customer support chatbot. AgentHub brings intelligent, personalized assistance to every step of the journey.

## A problem well-defined is half solved: Measuring Hypothesis Quality at Booking.com

DevFeed: [A problem well-defined is half solved: Measuring Hypothesis Quality at Booking.com](<https://devfeed.tech/articles/a-problem-well-defined-is-half-solved-measuring-hypothesis-quality-at-booking-com-61615.md>)

Original publisher: [Read original article](<https://booking.ai/a-problem-well-defined-is-half-solved-measuring-hypothesis-quality-at-booking-com-a85e845178a3?source=rss----4d265f07defc---4>)

Author: Luka Beverin

Published: 2026-09-29T09:42:29Z

Content type: article

Language: en

Sources: [Booking.com Data Science](<https://devfeed.tech/sources/booking-com-data-science.md>)

Topics: [product-experimentation](<https://devfeed.tech/topics/product-experimentation.md>), [A/B Testing](<https://devfeed.tech/topics/a-b-testing.md>), [AI-enhanced product development](<https://devfeed.tech/topics/ai-enhanced-product-development.md>), [Tool](<https://devfeed.tech/topics/tool.md>)

Tags: [adoption](<https://devfeed.tech/tags/adoption.md>), [ai](<https://devfeed.tech/tags/ai.md>), [ambassadors](<https://devfeed.tech/tags/ambassadors.md>), [analysis](<https://devfeed.tech/tags/analysis.md>), [behaviour](<https://devfeed.tech/tags/behaviour.md>), [booking](<https://devfeed.tech/tags/booking.md>), [change](<https://devfeed.tech/tags/change.md>), [experimentation](<https://devfeed.tech/tags/experimentation.md>), [featured](<https://devfeed.tech/tags/featured.md>), [hypothesis-testing](<https://devfeed.tech/tags/hypothesis-testing.md>), [implementation](<https://devfeed.tech/tags/implementation.md>), [measuring](<https://devfeed.tech/tags/measuring.md>), [metrics](<https://devfeed.tech/tags/metrics.md>), [problems](<https://devfeed.tech/tags/problems.md>), [quality](<https://devfeed.tech/tags/quality.md>), [reasoning](<https://devfeed.tech/tags/reasoning.md>), [records](<https://devfeed.tech/tags/records.md>), [reuse](<https://devfeed.tech/tags/reuse.md>), [template](<https://devfeed.tech/tags/template.md>), [test](<https://devfeed.tech/tags/test.md>), [theory](<https://devfeed.tech/tags/theory.md>), [tooling](<https://devfeed.tech/tags/tooling.md>)

### AI overview

Booking.com's advisory Hypothesis Quality Score uses an LLM to check whether experiment hypotheses state prior evidence, the change, target audience, expected user response and measurable effects. The in-house platform provides feedback during experiment creation, while a Quality Tab and product-track ambassadors help teams improve drafts. After rollout, the share of product improvement experiments rated Excellent rose from 62% to almost 90%; expected responses and observable metrics each gained 14 and 13 percentage points. The score measures how completely reasoning is documented, not idea value, experiment design or decision quality, and the reported gains do not prove the checker alone improved decisions. Booking.com keeps it advisory, pairs it with education and human review, and avoids automatic rewrites to encourage teams to clarify their own reasoning.

### Source excerpt

Even after an experiment (or A/B test) ends, it may not be clear what action to take or what was learned. The test may be properly powered, run for the right amount of time, and produce a statistically clear result. But if the hypothesis never made the original problem, reasoning for the change, or decision criteria explicit, the result is difficult to interpret. This is why hypothesis quality matters. A well-written hypothesis does more than improve an experiment at launch; it improves the quality of the decision that follows. By recording the evidence behind an idea, the user response we expect, and the metrics that would support or challenge it, teams create a clearer standard for interpreting the outcome. The benefits compound over time. A hypothesis that records only an implementation is difficult for another team (or an AI system), to find, compare, or reuse. A hypothesis that captures the underlying theory and expected behavioural response becomes a searchable explanation of why an experiment was run. Over time, this gives Booking.com a richer map of what we have learned: which user problems recur, which mechanisms tend to work, and where our assumptions repeatedly fail. In our earlier posts on experimentation quality and scaling that quality across Booking.com, we described how shared standards, measurement, education, and tooling can make better experimentation easier to practice at scale. This post focuses on the next layer of that work: how we are helping teams make the reasoning inside their hypotheses clearer, more consistent, and more useful long after an experiment has ended. The Hypothesis Template A hypothesis is not just a description of the feature we are about to build. It is the bridge between an observation, a theory about user behaviour, and the evidence that could support or challenge that theory. When that bridge is weak, the experiment may still run correctly, even score highly on other pillars of experimentation quality, but the result bec

## How we selected the next vector database at Booking.com

DevFeed: [How we selected the next vector database at Booking.com](<https://devfeed.tech/articles/how-we-selected-the-next-vector-database-at-booking-com-30452.md>)

Original publisher: [Read original article](<https://booking.ai/how-we-selected-the-next-vector-database-at-booking-com-1e738a5e3bb0?source=rss----4d265f07defc---4>)

Author: Başak Tuğçe Eskili

Published: 2026-08-11T10:31:50Z

Content type: article

Language: en

Sources: [Booking.com Data Science](<https://devfeed.tech/sources/booking-com-data-science.md>)

Topics: [Embeddings](<https://devfeed.tech/topics/embeddings.md>), [Vectors & LLMs](<https://devfeed.tech/topics/vectors-llms.md>), [vector embeddings in AI](<https://devfeed.tech/topics/vector-embeddings-in-ai.md>), [ann](<https://devfeed.tech/topics/ann.md>), [database-optimization](<https://devfeed.tech/topics/database-optimization.md>), [database-scalability](<https://devfeed.tech/topics/database-scalability.md>), [Artificial Intelligence](<https://devfeed.tech/topics/ai.md>), [Machine learning](<https://devfeed.tech/topics/machine-learning.md>)

Tags: [ai](<https://devfeed.tech/tags/ai.md>), [booking](<https://devfeed.tech/tags/booking.md>), [database](<https://devfeed.tech/tags/database.md>), [embeddings](<https://devfeed.tech/tags/embeddings.md>), [featured](<https://devfeed.tech/tags/featured.md>), [genai](<https://devfeed.tech/tags/genai.md>), [infrastructure](<https://devfeed.tech/tags/infrastructure.md>), [machine-learning](<https://devfeed.tech/tags/machine-learning.md>), [personalized-experiences](<https://devfeed.tech/tags/personalized-experiences.md>), [platform](<https://devfeed.tech/tags/platform.md>), [rag](<https://devfeed.tech/tags/rag.md>), [retrieval-augmented-generation](<https://devfeed.tech/tags/retrieval-augmented-generation.md>), [semantic](<https://devfeed.tech/tags/semantic.md>), [similarity-search](<https://devfeed.tech/tags/similarity-search.md>), [vector](<https://devfeed.tech/tags/vector.md>), [vector-database](<https://devfeed.tech/tags/vector-database.md>)

### AI overview

Booking.com selected Weaviate to replace OpenSearch as the backend for its internal Embedding Service after vector workloads grew to hundreds of millions of embeddings and OpenSearch required larger clusters, continual tuning, and higher operating costs. To compare systems, the team built a 100-million-vector dataset from Amazon Reviews 2023 using all-MiniLM-L6-v2, included metadata, and tested full and filtered nearest-neighbour search plus mixed read/write traffic at increasing concurrency. They tuned approximate search to about 99% recall and measured latency, throughput, resource use, and stability, emphasizing P99 latency. In these tests, specialized vector databases performed more efficiently than OpenSearch; Weaviate was most consistent across scenarios, and the team reported roughly 40% lower compute and memory usage cost at comparable recall and SLO targets. An existing service abstraction let Booking.com run databases in parallel and switch backends transparently, making migration largely a configuration change for most teams. The authors stress that these results reflect their production-shaped workloads and may not generalize.

### Source excerpt

This work was done in collaboration with Klaus Schaefers. Over the past few years, embeddings and vector search have become an important capability in many of our machine learning and GenAI systems at Booking.com. We initially started with a handful of use cases and experiments, and later this capability has grown into shared infrastructure that powers similarity search, semantic filtering, and retrieval-augmented generation (RAG). We used to treat vector search as a backend implementation detail, but today it directly drives the user experience. The real win isn't only speed but also the context. Expanding the variety of domain data we can retrieve efficiently gives our system the depth of context it needs to deliver accurate, and personalized experiences across the platform. This makes selecting the underlying vector database an infrastructure decision similar to choosing a primary datastore or message queue. It has to be predictable and scalable. As more teams started using our vector store, we began seeing highly diverse functional and operational requirements across different use cases. Some teams needed advanced capabilities like hybrid search or multi-vector support, while others demanded larger vector capacities and higher RPS metrics. These architectural needs ultimately brought us to a point where we needed to reassess whether our current setup could support this next phase of growth. Context: how embeddings fit into our stack Embeddings are vectors: fixed-length arrays of numbers produced by a model to represent an item (text, image, etc.). Each vector can be seen as a point in a high-dimensional space, where distance (or similarity) between points approximates semantic relatedness. By searching for the nearest vectors to a query vector, we retrieve items that are semantically "similar". This simple mechanism enables a wide range of use cases for us due its ability to do semantic similarity search. RAG-based use cases are the most well known examples. Ano

## A Recap of the 2026 Experimentation Conference at Booking.com

DevFeed: [A Recap of the 2026 Experimentation Conference at Booking.com](<https://devfeed.tech/articles/a-recap-of-the-2026-experimentation-conference-at-booking-com-30447.md>)

Original publisher: [Read original article](<https://booking.ai/a-recap-of-the-2026-experimentation-conference-at-booking-com-f43d48698fcd?source=rss----4d265f07defc---4>)

Author: Mel JI Mueller

Published: 2026-07-16T08:18:03Z

Content type: news

Language: en

Sources: [Booking.com Data Science](<https://devfeed.tech/sources/booking-com-data-science.md>)

Topics: [product-experimentation](<https://devfeed.tech/topics/product-experimentation.md>), [AI Chat](<https://devfeed.tech/topics/ai-chat.md>), [Data analysis](<https://devfeed.tech/topics/data-analysis.md>), [knowledge-management](<https://devfeed.tech/topics/knowledge-management.md>), [decision-making](<https://devfeed.tech/topics/decision-making.md>), [progressive delivery](<https://devfeed.tech/topics/progressive-delivery.md>), [Developer Tools](<https://devfeed.tech/topics/developer-tools.md>)

Tags: [2026](<https://devfeed.tech/tags/2026.md>), [ai](<https://devfeed.tech/tags/ai.md>), [booking](<https://devfeed.tech/tags/booking.md>), [conference](<https://devfeed.tech/tags/conference.md>), [culture](<https://devfeed.tech/tags/culture.md>), [data-driven](<https://devfeed.tech/tags/data-driven.md>), [data-science](<https://devfeed.tech/tags/data-science.md>), [decision-making](<https://devfeed.tech/tags/decision-making.md>), [events](<https://devfeed.tech/tags/events.md>), [experimentation](<https://devfeed.tech/tags/experimentation.md>), [experiments](<https://devfeed.tech/tags/experiments.md>), [recap](<https://devfeed.tech/tags/recap.md>), [scaling](<https://devfeed.tech/tags/scaling.md>)

### AI overview

The 2026 Booking.com Experimentation Conference brought together practitioners from 49 companies running about 56,000 experiments annually. Attendee survey results pointed to scaling, coordination, platform tooling, and culture as key challenges. Talks argued that AI-driven development increases the need for reliable rollout and measurement infrastructure, and showed how structured experiment records can make past results searchable. Deliveroo's opinionated tooling embeds statistical checks, guardrails, and fixed metrics to support trustworthy decisions. Booking.com Flights reported that validated micro-conversion metrics reduced average experiment time from 34 to 14 days in a pilot, while retaining the final business metric as a guardrail and checking impact through retests or blockouts. Culture sessions emphasized storytelling, education, and leadership incentives. Across the examples, the recap presents trustworthy experimentation as a combination of tooling, organizational practice, and human judgment.

### Source excerpt

By Kevin Anderson, Angelica Goetzen, Jorden Lentze, and Melanie Mueller On May 18, 2026, we hosted the third annual Experimentation Conference at Booking.com on our Amsterdam campus. What started in 2024 as an experiment itself -- would large-scale experimentation practitioners come together to learn from each other? -- has grown into an event which brings together over 150 practitioners from 49 companies which run experiments at scale. About one third of attendees came back a second or third time. The room collectively ran 56,000 experiments per year. It's a unique crowd, and that's exactly the point. The day opened with sharing the results of the survey data we collected from the participating companies on the state of experimentation across the room, revealing some interesting findings: most teams operate a centre of excellence model, roughly a third release over 90% of features through controlled experiments, and the top challenges are scaling, coordination, platform tooling, and culture. These shared experiences helped shape the programme. We had three sessions, grouped by the three conference themes: AI and experimentation: AI-assisted analysis, no-code experimentation Quality / velocity tradeoff: High-quality vs high-speed experimentation Experimentation culture: Build organizational buy-in and data-driven decision-making Each session followed the same format: two talks, then a panel discussion on the same topic. We closed with nine parallel breakout groups for deeper conversation. Below is a recap of the key sessions. Read the recap of 2025 | Read the recap of 2024 Session 1: AI and experimentation The conference started off with the hot topic of AI in experimentation. AI is changing how we experiment and how we support experimenters. How Experimentation Protects Decisions in an AI-Written World -- Marcel Toben Marcel Toben, Head of Engineering at Zalando, opened with a provocation he'd recently heard from software engineers in Berlin: nobody on his team had wr

## From Traditional ML to AI Agents: How Booking.com Scales AI Observability With Arize AI

DevFeed: [From Traditional ML to AI Agents: How Booking.com Scales AI Observability With Arize AI](<https://devfeed.tech/articles/from-traditional-ml-to-ai-agents-how-booking-com-scales-ai-observability-with-arize-ai-30450.md>)

Original publisher: [Read original article](<https://booking.ai/from-traditional-ml-to-ai-agents-how-booking-com-scales-ai-observability-with-arize-ai-625ac3996c7e?source=rss----4d265f07defc---4>)

Author: Amir Bitaraf

Published: 2026-07-10T07:52:18Z

Content type: article

Language: en

Sources: [Booking.com Data Science](<https://devfeed.tech/sources/booking-com-data-science.md>)

Topics: [ai observability](<https://devfeed.tech/topics/ai-observability.md>), [LLM observability](<https://devfeed.tech/topics/llm-observability.md>), [observability](<https://devfeed.tech/topics/observability.md>), [observability pipelines](<https://devfeed.tech/topics/observability-pipelines.md>), [OpenTelemetry](<https://devfeed.tech/topics/opentelemetry.md>), [agentic workflows](<https://devfeed.tech/topics/agentic-workflows.md>), [Machine learning](<https://devfeed.tech/topics/machine-learning.md>), [AI Context Management](<https://devfeed.tech/topics/ai-context-management.md>), [AI for customer service automation](<https://devfeed.tech/topics/ai-for-customer-service-automation.md>)

Tags: [agentic](<https://devfeed.tech/tags/agentic.md>), [agentic-ai](<https://devfeed.tech/tags/agentic-ai.md>), [agents](<https://devfeed.tech/tags/agents.md>), [ai](<https://devfeed.tech/tags/ai.md>), [ai-agents](<https://devfeed.tech/tags/ai-agents.md>), [ai-native](<https://devfeed.tech/tags/ai-native.md>), [ai-native-observability](<https://devfeed.tech/tags/ai-native-observability.md>), [ai-observability](<https://devfeed.tech/tags/ai-observability.md>), [arize](<https://devfeed.tech/tags/arize.md>), [automated](<https://devfeed.tech/tags/automated.md>), [background](<https://devfeed.tech/tags/background.md>), [balancing](<https://devfeed.tech/tags/balancing.md>), [batch](<https://devfeed.tech/tags/batch.md>), [behaviour](<https://devfeed.tech/tags/behaviour.md>), [blog-posts](<https://devfeed.tech/tags/blog-posts.md>), [booking](<https://devfeed.tech/tags/booking.md>), [building](<https://devfeed.tech/tags/building.md>), [can](<https://devfeed.tech/tags/can.md>), [ml](<https://devfeed.tech/tags/ml.md>), [observability](<https://devfeed.tech/tags/observability.md>), [with](<https://devfeed.tech/tags/with.md>)

### AI overview

Booking.com uses Arize AI as a shared observability platform for traditional ML and agentic AI, with different telemetry paths for each. Agent workflows emit OpenTelemetry traces with OpenInference conventions for model calls, retrieval, tools, guardrails, and related context; a central collector samples traces, applies PII-aware redaction, and routes them onward. Traditional ML uses prepared tabular prediction logs with features, predictions, labels, and model metadata, excluding sensitive data and limiting fields or sampling when needed. Arize supports trace inspection, alerts and dashboards, evaluations, production-derived datasets, plus drift, cohort, and feature-importance analysis for ML. In one production case, a latency alert led the team to identify an experimental model running without a service tier; coordinating with the provider restored expected latency. In prototyping, elevated P90 token counts and lower later-turn evaluations exposed long attraction URLs accumulating in chatbot history, allowing a fix before production.

### Source excerpt

Building an AI-native observability stack for agentic AI and traditional ML at Booking.com Authors: Amir Bitaraf, Shahaf Veber Why AI Observability Matters at Booking.com At Booking.com, AI helps travellers and partners in every step of their journey, from how people discover destinations to the way we support them while they're on the road. Rather than a single flagship model, we rely on a large and growing collection of systems that each solve a specific problem at scale. To make this concrete, consider a few examples: Trip planning assistants that help travelers turn vague ideas ("somewhere warm in April with good hiking") into concrete, bookable itineraries. On-site helpers that turn property details, amenities, reviews, and options into plain-language guidance, so people can choose the right stay with confidence. Partner copilots that help accommodation partners and other suppliers respond to guest messages faster and more consistently, while still staying in control of the final reply. Ranking systems that decide which options to show first in search and recommendation to surfaces, balancing user relevance with experimentation needs. Fraud detection models that quietly protect customers and partners in the background by flagging suspicious activity before it turns into real harm. Each of these systems is built and iterated on by different teams, uses different data, and runs under different constraints such as real-time vs batch, strict latency budgets vs more relaxed ones, fully automated vs human-in-the-loop. As we scale this ecosystem, observability becomes a first-class requirement, not a nice-to-have as we need to: Know when something changes in the real world, a new travel pattern, a data quality issue, a misconfiguration and how that affects model behaviour and user experience. Detect regressions early: slower responses, more confusing answers, drops in relevance or conversion, or subtle shifts that only show up for specific geographies, devices, or use

## Outlier Handling at Scale in Experimentation

DevFeed: [Outlier Handling at Scale in Experimentation](<https://devfeed.tech/articles/outlier-handling-at-scale-in-experimentation-30453.md>)

Original publisher: [Read original article](<https://booking.ai/outlier-handling-at-scale-in-experimentation-a8bb140e1ab8?source=rss----4d265f07defc---4>)

Author: Margarida Moreira da Silva

Published: 2026-07-01T13:44:26Z

Content type: article

Language: en

Sources: [Booking.com Data Science](<https://devfeed.tech/sources/booking-com-data-science.md>)

Topics: [product-experimentation](<https://devfeed.tech/topics/product-experimentation.md>), [experiments](<https://devfeed.tech/topics/experiments.md>)

Tags: [agencies](<https://devfeed.tech/tags/agencies.md>), [ate](<https://devfeed.tech/tags/ate.md>), [automated](<https://devfeed.tech/tags/automated.md>), [booking](<https://devfeed.tech/tags/booking.md>), [business](<https://devfeed.tech/tags/business.md>), [changes](<https://devfeed.tech/tags/changes.md>), [customer](<https://devfeed.tech/tags/customer.md>), [customer-experience](<https://devfeed.tech/tags/customer-experience.md>), [data](<https://devfeed.tech/tags/data.md>), [distribution](<https://devfeed.tech/tags/distribution.md>), [distributions](<https://devfeed.tech/tags/distributions.md>), [effect](<https://devfeed.tech/tags/effect.md>), [experiment](<https://devfeed.tech/tags/experiment.md>), [experimentation](<https://devfeed.tech/tags/experimentation.md>), [experiments](<https://devfeed.tech/tags/experiments.md>), [fail](<https://devfeed.tech/tags/fail.md>), [features](<https://devfeed.tech/tags/features.md>), [handling](<https://devfeed.tech/tags/handling.md>), [outlier](<https://devfeed.tech/tags/outlier.md>), [outlier-detection](<https://devfeed.tech/tags/outlier-detection.md>), [scale](<https://devfeed.tech/tags/scale.md>)

### AI overview

At Booking.com's scale, extreme observations can inflate variance and make experiment results unreliable. The article uses permutation tests on simulated A/A experiments: under no effect, p-values should be uniform; skew signals distortion. In simulations, failures became more likely as outlier size and frequency rose, and the maximum-to-99.9th-percentile ratio helped summarize the risk. Winsorization--capping rather than removing values--restored test validity in simulated and real distributions. Relative percentile caps adapt to each metric; absolute business-based caps are easier to explain. To select thresholds, the authors tested candidates across multiple experiments, kept those that consistently passed A/A tests, and looked for meaningful standard-deviation reduction before the mean shifted substantially. The approach trades variance reduction against bias: capping can hide genuine treatment effects in the tail, and absolute thresholds may depend on experiment duration. The authors describe threshold choice as a context-dependent judgment and note Huber-loss estimators as more principled but more complex and less scalable.

### Source excerpt

At Booking.com, thousands of experiments run simultaneously across highly heterogeneous users, from individual travellers to large travel agencies. This means our experiment data regularly contains legitimate but extreme values. When these go unhandled, they distort the statistical conclusions we draw, leading us to scale ideas that don't create value, or to discard ones that do. So, we need outlier handling methods that are reliable, automated, and applicable across diverse metrics without manual intervention. The Problem When extreme values are present in experiment data, they can compromise the estimation of average treatment effects (ATE), leading to unreliable test results and reduced statistical power. Even a single observation can inflate variance enough to mask a real effect or produce a spurious one. In practice, this means we risk shipping changes that appear positive but are not, or killing promising features because noise masked their real effect. At Booking.com's scale, this increase in false conclusions quickly compounds into a meaningful impact on customer experience and business outcomes. A Diagnostic Tool: the Permutation Test One way to assess whether extreme values are distorting results is the permutation test. By permuting over experiment data, we generate hundreds of simulated AA experiments where we know the ground truth: there is no real effect. Plotting the resulting p-values, we expect a uniform distribution. If it instead looks skewed, the underlying data distribution is compromising the validity of results. Plot 1: P-value distributions from simulated A/A tests. Clean normally-distributed estimated effects produce a uniform distribution (left), while the presence of extreme outliers results in skewed p-values (right), indicating a distorted false positive rate.Simulation Evidence: What Drives Failure? We ran AA permutation tests across a range of simulated data distributions to understand when they fail (i.e. not show a uniform p-value di

## Scaling Experimentation Quality at Booking.com

DevFeed: [Scaling Experimentation Quality at Booking.com](<https://devfeed.tech/articles/scaling-experimentation-quality-at-booking-com-30454.md>)

Original publisher: [Read original article](<https://booking.ai/scaling-experimentation-quality-at-booking-com-726152ee4ef0?source=rss----4d265f07defc---4>)

Author: Edgar Cano

Published: 2026-03-24T11:44:21Z

Content type: article

Language: en

Sources: [Booking.com Data Science](<https://devfeed.tech/sources/booking-com-data-science.md>)

Topics: [product-experimentation](<https://devfeed.tech/topics/product-experimentation.md>), [A/B Testing](<https://devfeed.tech/topics/a-b-testing.md>), [software-development](<https://devfeed.tech/topics/software-development.md>), [decision-making](<https://devfeed.tech/topics/decision-making.md>)

Tags: [000](<https://devfeed.tech/tags/000.md>), [booking](<https://devfeed.tech/tags/booking.md>), [data-science](<https://devfeed.tech/tags/data-science.md>), [experimentation](<https://devfeed.tech/tags/experimentation.md>), [featured](<https://devfeed.tech/tags/featured.md>), [quality](<https://devfeed.tech/tags/quality.md>), [scaling](<https://devfeed.tech/tags/scaling.md>)

### AI overview

Booking.com scaled experimentation quality by combining education and team ownership with measurement and changes to its in-house ET platform. Its three-pillar metric tracks experiment design, execution, and decision quality; dashboards initially showed that 80% of experiments lacked a power calculation. Community reviews and embedded Experiment Ambassadors helped teams address issues, while ET added a quality tab with corrective guidance, a power menu showing runtime and detectable effect tradeoffs, optional automatic power calculations, and defaults that limit access to non-decision data. Automatic power calculations reduce control over exact runtime. The authors report a steady rise in high-quality experiments, with the largest improvement in design, particularly power calculations and pre-registration. Product areas now own formal quality targets.

### Source excerpt

Authors: Edgar Cano, Daisy Duursma, Nils Skotara, Melanie Mueller Figure 1. Three-pillar components of Booking.com's Experimentation Quality Experimentation is at the core of product development in Booking.com, powered by our in-house platform, "ET" (Experiment Tool). At any given moment, we run approximately 1,000 parallel experiments to evaluate product changes. These experiments or A/B tests allow teams to directly compare a new version of the website against the existing one, validating hypotheses about how specific changes impact important metrics. In our organization, these pitfalls became more evident as our experiment volume grew. We observed that experimenters might set an arbitrary "two-week" duration without thinking about sufficient power, or extend a test until results "became significant" or "trended positive." Knowing that this lack of consistency leads to flawed decision-making we dedicated significant effort to increasing the quality of our experimentation process, making Experimentation Quality a key KPI for our program. However, identifying the problem was only the start; the greater challenge is how to implement these standards across a large organization. Enforcement vs. Education When deciding how to scale quality, we faced a fundamental choice: Do we enforce strict controls or we rely on education. Ultimately, we left it to product teams to decide how to conduct their experiments. This choice entailed several trade-offs: Enforcement: Ensures comparability, consistency, and reliability. However, it comes at the cost of flexibility. There is a risk that people follow "rules" blindly without understanding the rationale. Education: Aims for a culture where experimenters understand the why behind best practices. This leads to better buy-in, allows teams to challenge methods, and highlights individual responsibility. It also prevents bottlenecking. If enforcement requires human review, it slows down development. However, education requires a massive

## Shape the Future of Travel: Join our 2026 GenAI & ML PhD Research Internship!

DevFeed: [Shape the Future of Travel: Join our 2026 GenAI & ML PhD Research Internship!](<https://devfeed.tech/articles/shape-the-future-of-travel-join-our-2026-genai-ml-phd-research-internship-30455.md>)

Original publisher: [Read original article](<https://booking.ai/shape-the-future-of-travel-join-our-2026-genai-ml-phd-research-internship-a36793c34fbc?source=rss----4d265f07defc---4>)

Author: Yang Yang

Published: 2026-02-05T10:39:26Z

Content type: news

Language: en

Sources: [Booking.com Data Science](<https://devfeed.tech/sources/booking-com-data-science.md>)

Topics: [Artificial Intelligence](<https://devfeed.tech/topics/ai.md>), [Machine learning](<https://devfeed.tech/topics/machine-learning.md>), [Generative AI](<https://devfeed.tech/topics/generative-ai.md>), [llm-reasoning](<https://devfeed.tech/topics/llm-reasoning.md>), [Embeddings](<https://devfeed.tech/topics/embeddings.md>), [big-data](<https://devfeed.tech/topics/big-data.md>), [Image](<https://devfeed.tech/topics/image.md>)

Tags: [2026](<https://devfeed.tech/tags/2026.md>), [3](<https://devfeed.tech/tags/3.md>), [agent](<https://devfeed.tech/tags/agent.md>), [ai](<https://devfeed.tech/tags/ai.md>), [big-data](<https://devfeed.tech/tags/big-data.md>), [blog-posts](<https://devfeed.tech/tags/blog-posts.md>), [booking](<https://devfeed.tech/tags/booking.md>), [challenges](<https://devfeed.tech/tags/challenges.md>), [collaboration](<https://devfeed.tech/tags/collaboration.md>), [community](<https://devfeed.tech/tags/community.md>), [complex](<https://devfeed.tech/tags/complex.md>), [context](<https://devfeed.tech/tags/context.md>), [context-engineering](<https://devfeed.tech/tags/context-engineering.md>), [featured](<https://devfeed.tech/tags/featured.md>), [future](<https://devfeed.tech/tags/future.md>), [genai](<https://devfeed.tech/tags/genai.md>), [internship](<https://devfeed.tech/tags/internship.md>), [join](<https://devfeed.tech/tags/join.md>), [machine-learning](<https://devfeed.tech/tags/machine-learning.md>), [ml](<https://devfeed.tech/tags/ml.md>), [phd-research](<https://devfeed.tech/tags/phd-research.md>), [research](<https://devfeed.tech/tags/research.md>), [travel](<https://devfeed.tech/tags/travel.md>)

### AI overview

Booking.com's 2026 PhD research internship offers a three-month placement with its Machine Learning teams in Amsterdam, working on real-world travel problems. Project areas include target encoding, multi-agent collaboration, reinforcement learning from user feedback, transformer explainability for sequential data, embeddings, retrieval for travel planning, and synthetic image data. Applicants should be current PhD students in quantitative fields with independent research ability, strong machine-learning knowledge, Python and big-data skills (SQL, Spark, or Hadoop), and peer-reviewed publications; individual projects may have additional requirements. The internship is in the Netherlands. Applications are due March 1, 2026, and should include a CV and the supervisor's email address for a later recommendation request.

### Source excerpt

At Booking.com, we don't just use Machine Learning -- we use it to solve some of the most complex travel challenges in the world. We're looking for the next generation of researchers to join our Machine Learning community in Amsterdam for a 3-month deep dive into cutting-edge AI. The Program As a Research Intern, you'll be embedded in our teams, working alongside world-class mentors. Your mission? To tackle real-world problems and push the boundaries of the state-of-the-art. Are You the One? We're looking for current PhD students in quantitative fields (CS, Math, AI, Physics) who can conduct independent research and have a solid grip on Python and Big Data tech (SQL, Spark, Hadoop). What's in it for you? You won't just be "an intern". You'll be a contributor to our Machine Learning community. You'll have the opportunity to contribute to the existing efforts of the Machine Learning teams, participate in internal knowledge-sharing sessions, and enjoy the collaborative, high-energy environment of our Amsterdam HQ. Projects Regularized Target Encoding for large real-world datasets Multi-Agent Collaboration Aligning LLMs with user feedback via reinforcement learning Multi-level treatments Interpretable Foundations: Explainability Methods for Transformer Models on Sequential Event Data Scalable and generalisable ID embedding learning Improving property embeddings with better handling of rich and long-context data Utility-aware retrieval for context engineering in travel planning Synthetic Data Generation in Images Requirements We are looking for independent researchers with strong understanding of Machine Learning topics (see requirements for each project in the Linkedin ad), have a track record of peer-reviewed publications and a passion for solving complex problems. Why Booking.com? You'll join a vibrant, diverse community of data scientists and researchers who love to experiment. Beyond the code, you'll experience the unique culture of our Amsterdam headquarters -- a hub

## Beyond Prompt Engineering: How We Used Supervised Fine-Tuning for Travel Recommendations

DevFeed: [Beyond Prompt Engineering: How We Used Supervised Fine-Tuning for Travel Recommendations](<https://devfeed.tech/articles/beyond-prompt-engineering-how-we-used-supervised-fine-tuning-for-travel-recommendations-30449.md>)

Original publisher: [Read original article](<https://booking.ai/beyond-prompt-engineering-how-we-used-supervised-fine-tuning-for-travel-recommendations-91e8f4711e4b?source=rss----4d265f07defc---4>)

Author: Amit Meitin

Published: 2026-01-29T10:53:21Z

Content type: article

Language: en

Sources: [Booking.com Data Science](<https://devfeed.tech/sources/booking-com-data-science.md>)

Topics: [Fine-tuning](<https://devfeed.tech/topics/fine-tuning.md>), [recommendation systems](<https://devfeed.tech/topics/recommendation-systems.md>), [Large Language Model](<https://devfeed.tech/topics/llm.md>), [AI Chat](<https://devfeed.tech/topics/ai-chat.md>), [Enhancing AI model performance](<https://devfeed.tech/topics/enhancing-ai-model-performance.md>), [Machine learning](<https://devfeed.tech/topics/machine-learning.md>), [Conversational AI](<https://devfeed.tech/topics/conversational-ai.md>), [LLM optimization](<https://devfeed.tech/topics/llm-optimization.md>), [Model Development](<https://devfeed.tech/topics/model-development.md>)

Tags: [ai](<https://devfeed.tech/tags/ai.md>), [app](<https://devfeed.tech/tags/app.md>), [blog-posts](<https://devfeed.tech/tags/blog-posts.md>), [booking](<https://devfeed.tech/tags/booking.md>), [chat](<https://devfeed.tech/tags/chat.md>), [clicks](<https://devfeed.tech/tags/clicks.md>), [data](<https://devfeed.tech/tags/data.md>), [fine-tuning](<https://devfeed.tech/tags/fine-tuning.md>), [inference](<https://devfeed.tech/tags/inference.md>), [language-models](<https://devfeed.tech/tags/language-models.md>), [languages](<https://devfeed.tech/tags/languages.md>), [large-language-models](<https://devfeed.tech/tags/large-language-models.md>), [large-language-models-llms](<https://devfeed.tech/tags/large-language-models-llms.md>), [llms](<https://devfeed.tech/tags/llms.md>), [machine-learning](<https://devfeed.tech/tags/machine-learning.md>), [metrics](<https://devfeed.tech/tags/metrics.md>), [prompt-engineering](<https://devfeed.tech/tags/prompt-engineering.md>), [real-time](<https://devfeed.tech/tags/real-time.md>), [recommendation](<https://devfeed.tech/tags/recommendation.md>), [recommendations](<https://devfeed.tech/tags/recommendations.md>), [travel](<https://devfeed.tech/tags/travel.md>)

### AI overview

Booking.com reports that a supervised fine-tuned, 7B open-weight model improved destination recommendations and reduced p99 inference latency by 67% (about 3x) versus its production hybrid of an API-based proprietary LLM and traditional ML. It used LoRA/QLoRA adapters with frozen base weights, trained on conversational inputs and ordered destination labels derived from bookings, clicks, and filtered views. Adding user location improved Hit@5 by 8% in the reported example evaluation; prepending the desired recommendation count helped the model return exactly K items. The excerpt says offline gains were later validated in an A/B test, but omits the offline metrics table and provides no A/B results. The system is hosted internally, allowing use of sensitive context without sending it to an external LLM provider.

### Source excerpt

How fine-tuning delivered superior recommendation metrics while achieving 3x faster inference Every traveler is unique, and so is every trip. At Booking.com, we're always looking for new ways to make trip planning feel less like a chore and more like an adventure. That's why we set out to build an AI Trip Planner that doesn't just answer questions, but actually understands what travelers want, expressed in ways they are most comfortable with. But here's the challenge: travelers today are expressing their needs in ways that are more unstructured and nuanced than ever before. Instead of ticking boxes or picking from drop-down menus, people describe their dream vacations in their own words. Sometimes these requests are vague, sometimes very specific, but always personal. Traditional machine learning models, which thrive on structured data and clear signals, can struggle to keep up with this new level of expressiveness. At the same time, we at Booking.com have years of valuable, structured data from searches, clicks, and bookings. This data captures what travelers actually do: the trips they plan and the vacation they go on. The big question for us became how to combine the best of both worlds -- How can we harness the power of Large Language Models (LLMs) to understand and respond to unstructured requests, and use the wealth of behavioral data that has driven our recommendations for years? This post is about how we tackled that question and what we learned along the way. What is the AI Trip Planner? The AI Trip Planner is Booking.com's conversational assistant that helps travelers find destinations, accommodations and other trip components through natural, real-time chat. Instead of searching with filters, users describe what they want in their own words. The planner suggests options, whether it is inspiration or specific places to stay. Integrated into the app, it makes planning and discovering seamless and is now available in multiple countries and languages. Bridging

## AI Agent Evaluation

DevFeed: [AI Agent Evaluation](<https://devfeed.tech/articles/ai-agent-evaluation-30448.md>)

Original publisher: [Read original article](<https://booking.ai/ai-agent-evaluation-82e781439d97?source=rss----4d265f07defc---4>)

Author: Antonio Castelli

Published: 2026-01-21T13:11:23Z

Content type: tutorial

Language: en

Sources: [Booking.com Data Science](<https://devfeed.tech/sources/booking-com-data-science.md>)

Topics: [LLM evaluation / benchmarking](<https://devfeed.tech/topics/llm-evaluation-benchmarking.md>), [AI Agent](<https://devfeed.tech/topics/ai-agent.md>), [Artificial Intelligence](<https://devfeed.tech/topics/ai.md>), [Tool](<https://devfeed.tech/topics/tool.md>)

Tags: [agent](<https://devfeed.tech/tags/agent.md>), [ai](<https://devfeed.tech/tags/ai.md>), [ai-agent](<https://devfeed.tech/tags/ai-agent.md>), [api](<https://devfeed.tech/tags/api.md>), [best-practices](<https://devfeed.tech/tags/best-practices.md>), [blog-posts](<https://devfeed.tech/tags/blog-posts.md>), [evaluation](<https://devfeed.tech/tags/evaluation.md>), [llm-agents](<https://devfeed.tech/tags/llm-agents.md>), [llm-evaluation](<https://devfeed.tech/tags/llm-evaluation.md>), [machine-learning](<https://devfeed.tech/tags/machine-learning.md>), [sql](<https://devfeed.tech/tags/sql.md>), [tools](<https://devfeed.tech/tags/tools.md>)

### AI overview

Booking.com's article presents practical guidance for evaluating LLM agents. It explains how agents differ from single LLMs because they can use external tools and iterate through thoughts, actions, and observations, then introduces black-box and glass-box evaluation approaches.

### Source excerpt

AI Agent Evaluation: practical tips at Booking.com Authors: Zeno Belligoli, Antonio Castelli, George Chouliaras This article is the 2nd part of our GenAI evaluation best practices series. You can read the first part, focusing on single LLM evaluation, here. 1. Overview LLM agents are advanced AI systems that leverage large language models to perform complex tasks autonomously. Unlike single LLMs that respond to direct prompts, agents can break down problems, use external tools, and iterate on solutions through a series of thoughts, actions, and observations. This allows them to achieve goals that require more intricate planning, reasoning, and interaction with various environments. Agents can autonomously determine if tool utilization is necessary to fulfill a user's request. For example, they might employ: a calculator for mathematical operations, a flights API to retrieve available flights for a certain destination, or execute a SQL query to fetch information about an hotel reservation from a database. Figure1: Schematic view of an agent. The user might interact with the agent within a multi-turn conversation, asking it to perform various tasks (recommend vacation destinations, ask questions about a hotel etc.). The agent should try to complete the requested tasks abiding by a set of rules (or constraints) provided by the product specification (e.g. do not recommend properties outside the inventory). In performing the tasks the agent has freedom to use a certain set of tools, and it can interact with them via API requests. Tool examples typically range from general (e.g. calculate travel distance) to use case specific functionalities (e.g. retrieving information from a database). While these enhanced capabilities are fundamental to solve tasks which would be difficult (or even impossible) to solve with text generation only, they require a more complex evaluation process compared to evaluating a single "prompt-response" LLM. The evaluation of an agent's performance

## Highlights of Booking.com's publication in 2025

DevFeed: [Highlights of Booking.com's publication in 2025](<https://devfeed.tech/articles/highlights-of-booking-com-s-publication-in-2025-30451.md>)

Original publisher: [Read original article](<https://booking.ai/highlights-of-booking-coms-publication-in-2025-1c1a6deba066?source=rss----4d265f07defc---4>)

Author: Yang Yang

Published: 2026-01-20T09:20:13Z

Content type: article

Language: en

Sources: [Booking.com Data Science](<https://devfeed.tech/sources/booking-com-data-science.md>)

Topics: [Machine learning](<https://devfeed.tech/topics/machine-learning.md>), [Artificial Intelligence](<https://devfeed.tech/topics/ai.md>), [Natural language processing](<https://devfeed.tech/topics/nlp.md>), [Language models](<https://devfeed.tech/topics/language-models.md>), [Fine-tuning](<https://devfeed.tech/topics/fine-tuning.md>), [recommendation systems](<https://devfeed.tech/topics/recommendation-systems.md>), [Optimization](<https://devfeed.tech/topics/optimization.md>), [Compression](<https://devfeed.tech/topics/compression.md>), [Inference](<https://devfeed.tech/topics/inference.md>)

Tags: [2025](<https://devfeed.tech/tags/2025.md>), [ai](<https://devfeed.tech/tags/ai.md>), [booking](<https://devfeed.tech/tags/booking.md>), [compression](<https://devfeed.tech/tags/compression.md>), [data-science](<https://devfeed.tech/tags/data-science.md>), [fine-tuning](<https://devfeed.tech/tags/fine-tuning.md>), [inference](<https://devfeed.tech/tags/inference.md>), [language-models](<https://devfeed.tech/tags/language-models.md>), [latency](<https://devfeed.tech/tags/latency.md>), [neurips](<https://devfeed.tech/tags/neurips.md>), [nlp](<https://devfeed.tech/tags/nlp.md>), [optimization](<https://devfeed.tech/tags/optimization.md>), [publication](<https://devfeed.tech/tags/publication.md>)

### AI overview

Booking.com highlights its 2025 machine learning publications, including papers accepted at major conferences and research on applying Medusa speculative decoding and knowledge distillation to travel-related language model tasks.

### Source excerpt

At Booking.com, our mission is to make experiencing the world easier for everyone. We are committed to investing in cutting-edge technology that removes the barriers to travel, enabling seamless connections between millions of travelers and unforgettable experiences, diverse transportation options, and exceptional accommodations. The intersection of academic rigor and industry application is where true transformation happens. In 2025, our ML community bridged this gap more effectively than ever, contributing vital new insights to the global scientific community. With 8 out of 13 papers accepted at premier conferences -- including NeurIPS, SIGIR, KDD, and ACL -- our colleagues have demonstrated world-class expertise in AI, NLP, recommendation systems, uplift modeling, etc. These aren't just theoretical wins; they are the engines of innovation that allow us to push technological boundaries, ensuring our platform remains the most sophisticated and intuitive guide in the ever-evolving travel industry. Below, we highlight some of the key achievements and insights from these groundbreaking works. Speed Without Sacrifice: Fine-Tuning Language Models with Medusa and Knowledge Distillation in Travel Applications By Daniel Zagyva, Emmanouil Stergiadis, Laurens Van Der Maas, Aleksandra Dokic, Eran Fainman, Ilya Gusev, Moran Beladev Best paper award of 2025 ACL Industry Track https://aclanthology.org/2025.acl-industry.48/ In high-stakes industrial NLP applications, balancing generation quality with speed and efficiency presents significant challenges. We address them by investigating two complementary optimization approaches: Medusa for speculative decoding and knowledge distillation (KD) for model compression. We demonstrate the practical application of these techniques in real-world travel domain tasks, including trip planning, smart filters, and generating accommodation descriptions. We introduce modifications to the Medusa implementation, starting with base pre-trained models