# Pydantic Python data validation library that uses type annotations for schema validation and serialization. This is one page of public article previews, not the complete archive. Follow Next page to continue. Summaries are not the original full articles. ## Write End-to-End Tests in Your Backend's Language DevFeed: [Write End-to-End Tests in Your Backend's Language]() Original publisher: [Read original article]() Author: James McConkey Published: 2026-09-17T12:00:42Z Content type: tutorial Language: en Sources: [Atomic Object]() Topics: [Testing](), [Playwright](), [FastAPI](), [Python](), [test data](), [Pydantic](), [SQLAlchemy](), [ASP.NET Core](), [PostgreSQL](), [Docker Compose]() Tags: [asp-net-core](), [code](), [data](), [docker-compose](), [fastapi](), [playwright](), [postgresql](), [project-team-management](), [python](), [sqlalchemy](), [tests](), [the-software-life]() ### AI overview This article argues that end-to-end test-data setup is often the main design challenge, because tests must create consistent domain records while running alongside other tests. It recommends using browser-testing tools in the backend's language when possible, keeping meaningful relationships inline, and extracting small creation helpers without hiding scenario intent. ### Source excerpt The browser is often the easiest part of an end-to-end test. Consider a test that verifies a user can complete an overdue task. The visible interaction is small: sign in, find the task, click Complete, and observe the new status. Before any of that can happen, the test needs a workspace, a user, a project, and [...] The post Write End-to-End Tests in Your Backend's Language appeared first on Atomic Spin. ## Browser automation with Pydantic AI + Playwright DevFeed: [Browser automation with Pydantic AI + Playwright]() Original publisher: [Read original article]() Author: Pamela Fox (noreply@blogger.com) Published: 2026-08-20T21:57:18Z Content type: tutorial Language: en Sources: [Pamela Fox]() Topics: [Playwright](), [Pydantic](), [Browser Automation](), [Artificial Intelligence](), [Azure OpenAI](), [Entra ID](), [Authentication](), [OAuth 2.0](), [OpenTelemetry]() Tags: [agents](), [ai](), [api-keys](), [authentication](), [automation](), [azure-openai](), [browser](), [microsoft](), [microsoft-foundry](), [openai](), [playwright](), [python]() ### AI overview This tutorial explains how to combine Pydantic AI with Playwright to build agents that browse websites programmatically. It covers connecting Pydantic AI to Microsoft Foundry models through an OpenAI-compatible endpoint, using Entra token-based authentication, and applying Playwright for browsing, manual QA, and design iteration. ### Source excerpt When we build agents, we often want to give them the ability to browse the web: open webpages, navigate from one page to the other, and read the content of a webpage. By combining Pydantic AI with the Playwright capability from Pydantic AI Harness, we can build agents that browse the web safely and programmatically. Using Pydantic AI with Microsoft Foundry models Pydantic AI is an open-source model-agnostic framework from Pydantic for building LLM-based applications and agents. It's type-safe and supports OpenTelemetry, making it a great choice for robust production applications. We can use Pydantic-AI with Microsoft Foundry models using either API keys or Entra token-based authentication. When possible, we always recommend the keyless route, so that's what we'll demonstrate here. We use the azure-identity package to authenticate with Entra, using either local or managed identity, and get back a token provider callback function for that credential: from azure.identity.aio import AzureDeveloperCliCredential, get_bearer_token_provider credential = AzureDeveloperCliCredential() token_provider = get_bearer_token_provider(credential, "https://cognitiveservices.azure.com/.default") Then we use the OpenAI package to configure the model connection: from openai import AsyncOpenAI client = AsyncOpenAI( base_url=os.environ["AZURE_OPENAI_ENDPOINT"] + "/openai/v1", api_key=token_provider, ) model = OpenAIChatModel( model_name=os.environ["AZURE_OPENAI_CHAT_DEPLOYMENT"], provider=OpenAIProvider(openai_client=client), ) Let's explain the options used above: base_url: We point this at the OpenAI-compatible endpoint for our Foundry model. This endpoint works for Azure OpenAI models (like gpt-5.4, which this project deploys), and for cross-provider Foundry models that support the OpenAI v1 API, like Kimi-K2.7-Code. The base URL looks like "https://AZURE_OPENAI_SERVICE_NAME.openai.azure.com/openai/v1". api_key: We pass in the token provider callback function that generates OAuth2 token ## Learning Path: FastAPI: Python API Development With Light Speed DevFeed: [Learning Path: FastAPI: Python API Development With Light Speed]() Original publisher: [Read original article]() Author: Real Python Published: 2026-07-22T12:00:00Z Content type: tutorial Language: en Sources: [Real Python]() Topics: [FastAPI](), [Python](), [API](), [REST API](), [Pydantic](), [Web Development](), [CRUD]() Tags: [api](), [api-documentation](), [database](), [development](), [python](), [rest](), [tutorial](), [validation](), [web-development]() ### AI overview This learning path teaches Python developers how to build REST APIs and complete web applications with FastAPI. It covers automatic validation, serialization, interactive documentation, Pydantic data models, CRUD operations, Jinja2-rendered web pages, and a URL shortener project. ### Source excerpt Learn FastAPI from the ground up. Build REST APIs, serve web pages with Jinja2 templates, and create a complete URL shortener project in Python. ## Starlette vs FastAPI: what FastAPI actually adds DevFeed: [Starlette vs FastAPI: what FastAPI actually adds]() Original publisher: [Read original article]() Author: Farhan Hasin Chowdhury Published: 2026-07-20T07:00:00Z Content type: comparison Language: en Sources: [Honeybadger]() Topics: [FastAPI](), [Pydantic](), [ASGI](), [OpenAPI Specification](), [Dependency injection](), [Python](), [WebSocket](), [Cross-origin resource sharing (CORS)]() Tags: [cors](), [data-validation](), [dependency-injection](), [fastapi](), [frameworks](), [openapi](), [python](), [python-articles](), [starlette](), [websocket]() ### AI overview This comparison explains how FastAPI builds on Starlette and Pydantic. Starlette provides the ASGI-based HTTP layer, while Pydantic handles typed data validation; FastAPI adds type-driven parameter parsing, dependency injection, and automatic OpenAPI documentation. It also discusses when using raw Starlette may be preferable. ### Source excerpt FastAPI is built on Starlette, but most developers never look at what's underneath. Learn what FastAPI actually adds on top of Starlette and Pydantic, what comes straight from Starlette, and when dropping down to raw Starlette makes more sense than pulling in the full stack. ## Using Evaluation Frameworks with Agent Observability DevFeed: [Using Evaluation Frameworks with Agent Observability]() Original publisher: [Read original article]() Author: Jennifer Mickel; Eddie Cai Published: 2026-06-22T00:00:00Z Content type: article Language: en Sources: [Datadog | The Monitor blog]() Topics: [agent observability](), [observability](), [Large Language Model](), [Pydantic](), [experiments](), [Traces](), [Retrieval Augmented Generation (RAG)]() Tags: [agent-observability](), [ai-observability](), [code](), [deployment](), [development](), [evals](), [integration](), [llm](), [observability](), [traces]() ### AI overview This article explains how Datadog Agent Observability integrates existing DeepEval and Pydantic Evals frameworks. It covers running evaluations in experiments, connecting evaluation scores to production traces, and continuously monitoring LLM evaluation quality across development and deployment. ### Source excerpt Run DeepEval and Pydantic Evals natively in Datadog Agent Observability. Track regressions and connect eval scores to production traces. ## From Contributor to Outreachy Intern: My Fedora Journey Begins DevFeed: [From Contributor to Outreachy Intern: My Fedora Journey Begins]() Original publisher: [Read original article]() Author: Aman . Published: 2026-06-16T12:00:00Z Content type: article Language: en Sources: [Fedora Community Blog]() Topics: [Fedora](), [FastAPI](), [API](), [release engineering](), [Authentication](), [CI/CD](), [forgejo](), [Pydantic](), [Swagger](), [Deployment](), [openid]() Tags: [apis](), [backend](), [blog](), [fedora-project-community](), [intern](), [journey](), [mentored-projects](), [open-source](), [outreachy]() ### AI overview Aman describes his first two weeks as an Outreachy intern with the Fedora community. He introduces the Fedora Release Schedule Planner API project and outlines planned work on testing, CI reliability, OpenID Connect authentication, Fedora infrastructure integration, and deployment preparation. ### Source excerpt Introduction Hi, I'm Aman, a software developer and open-source enthusiast who enjoys backend development, APIs, and building tools that are useful to others. I recently started my Outreachy internship with the Fedora community, and this is my first blog as an intern. In this post, I want to share what my first two weeks have [...] The post From Contributor to Outreachy Intern: My Fedora Journey Begins appeared first on Fedora Community Blog. ## How to build deep research agents using Temporal and Braintrust DevFeed: [How to build deep research agents using Temporal and Braintrust]() Original publisher: [Read original article]() Author: Martin Bergman Published: 2026-06-03T00:00:00Z Content type: tutorial Language: en Sources: [Temporal Blog]() Topics: [Multi Agent Systems](), [observability](), [Pydantic](), [parallel](), [retry](), [Large Language Model]() Tags: [agents](), [build](), [how-to](), [multi-agent](), [observability](), [parallel](), [research](), [retry](), [tracing]() ### AI overview This tutorial explains a multi-agent deep research pipeline built with Temporal and Braintrust. It covers planning, query generation, parallel web search, report synthesis, and the use of Durable Execution, evals, and observability to handle timeouts, partial failures, and difficult debugging. ### Source excerpt Deep research agents are fragile in production. Here's how Temporal and Braintrust make them resilient with Durable Execution, evals, and tracing. ## Same-Day Model Integration: ESMC + ESMFold2 in Sheaf v0.11 DevFeed: [Same-Day Model Integration: ESMC + ESMFold2 in Sheaf v0.11]() Original publisher: [Read original article]() Published: 2026-05-27T00:00:00Z Content type: article Language: en Sources: [Alex Korbonits]() Topics: [Large Language Model](), [structure](), [Pydantic](), [backends](), [unit tests](), [Serverless]() Tags: [backend](), [claude-code](), [code](), [language](), [model](), [pypi](), [release](), [serverless](), [structure](), [tests]() ### AI overview The article explains how Sheaf v0.11 integrated ESMC and ESMFold2 on the same day they were released. It attributes the rapid integration to existing typed serving contracts and infrastructure, while describing verification, testing, H100 smoke testing, release work, and documentation. ### Source excerpt Chan Zuckerberg Biohub released a new protein language model and structure predictor this morning. Sheaf v0.11 shipped with both, same day. The story is less about hustle and more about what a typed serving contract buys you when a new model lands. ## Introducing the Common AI Provider: LLM and AI Agent Support for Apache Airflow DevFeed: [Introducing the Common AI Provider: LLM and AI Agent Support for Apache Airflow]() Original publisher: [Read original article]() Author: Apache Airflow Published: 2026-04-14T00:00:00Z Content type: release Language: en Sources: [Apache Airflow Blog]() Topics: [airflow](), [Large Language Model](), [AI Agent](), [Pydantic](), [Artificial Intelligence]() Tags: [2025](), [ai](), [ai-agent](), [airflow](), [apache](), [apache-airflow](), [community](), [llm](), [release]() ### AI overview Apache Airflow is releasing the Common AI Provider 0.1.0, a provider package that adds LLM and AI agent capabilities directly to Airflow. Built on Pydantic AI, it supports more than 20 model providers through a single install and requires Apache Airflow 3.0 or newer. ### Source excerpt At Airflow Summit 2025, we previewed what native AI integration in Apache Airflow could look like. Today we're shipping it. apache-airflow-providers-common-ai 0.1.0 adds LLM and agent capabilities directly to Airflow. Not a wrapper around another framework, but a provider package that plugs into the orchestrator you already run. It's built on Pydantic AI and supports 20+ model providers (OpenAI, Anthropic, Google, Azure, Bedrock, Ollama, and more) through a single install. pip install 'apache-airflow-providers-common-ai' Requires Apache Airflow 3.0+. Note: This is a 0.x release. We're actively looking for feedback and iterating fast, so breaking changes are possible between minor versions. Try it, tell us what works and what doesn't. Your input directly shapes the API. By the Numbers 6 Operators 6 TaskFlow decorators 5 Toolsets 4 Connection types 20+ Supported model providers via Pydantic AI The Decorator Suite Every operator has a matching TaskFlow decorator. @task.llm: Single LLM Call Send a prompt, get text or structured output back. from pydantic import BaseModel from airflow.providers.common.compat.sdk import dag, task @dag def my_pipeline(): class Entities(BaseModel): names: list[str] locations: list[str] @task.llm( llm_conn_id="my_openai_conn", system_prompt="Extract named entities.", output_type=Entities, ) def extract(text: str): return f"Extract entities from: {text}" extract("Alice visited Paris and met Bob in London.") my_pipeline() The LLM returns a typed Entities object, not a string you have to parse. Downstream tasks get structured data through XCom. @task.agent: Multi-Step Agent with Tools When the LLM needs to query databases, call APIs, or read files across multiple steps, use @task.agent. The agent picks which tools to call and loops until it has an answer. from airflow.providers.common.ai.toolsets.sql import SQLToolset from airflow.providers.common.compat.sdk import dag, task @dag def sql_analyst(): @task.agent( llm_conn_id="my_openai_conn", sys ## How to Accept Payments in a FastAPI Backend DevFeed: [How to Accept Payments in a FastAPI Backend]() Original publisher: [Read original article]() Author: Ayush Agarwal Published: 2026-03-29T00:00:00Z Content type: tutorial Language: en Sources: [Dodo Payments Blog]() Topics: [FastAPI](), [Back end](), [Pydantic](), [Python](), [SDKs](), [API](), [Software as a service]() Tags: [api](), [async](), [backend](), [checkout](), [compliance](), [developer-tools](), [fastapi](), [how-to](), [merchant-of-record](), [payment-gateway](), [payments](), [python](), [sdk](), [swagger](), [tax](), [validation](), [verification](), [webhooks]() ### AI overview A tutorial on integrating Dodo Payments into a FastAPI backend. It covers asynchronous checkout sessions and webhooks, Pydantic request validation, signature verification, and using verified webhook events as the source of truth for payment state. ### Source excerpt Learn how to integrate Dodo Payments into your FastAPI backend with async webhook handlers, Pydantic validation, and secure signature verification. ## Structured outputs with Pydantic AI DevFeed: [Structured outputs with Pydantic AI]() Original publisher: [Read original article]() Author: Ed Berry Published: 2026-03-24T10:34:08Z Content type: tutorial Language: en Sources: [FreeAgent]() Topics: [Large Language Model](), [JSON Schema](), [Prompt Engineering](), [Pydantic](), [Python](), [Structured-data]() Tags: [ai](), [data](), [data-ml](), [data-science](), [data-validation](), [genai](), [json-schema](), [llms](), [prompt-engineering](), [python](), [schema](), [structured-output]() ### AI overview This article explains structured outputs for LLMs using Pydantic AI. It introduces the relationship between Python type hints, static analysis, Pydantic, and runtime data validation, while noting that Pydantic AI is model-agnostic. ### Source excerpt One of the challenges of working with LLMs is getting them to respond with a consistent format, such as a given JSON schema. Anyone who has tried to solve this issue with prompt engineering knows how frustrating it can be. You add a 'MUST' here and an 'always return JSON' there, but still the output [...] ## Building a Local AI Task Manager with PydanticAI and Ollama DevFeed: [Building a Local AI Task Manager with PydanticAI and Ollama]() Original publisher: [Read original article]() Author: Alex Razvant Published: 2026-03-22T14:03:18Z Content type: tutorial Language: en Sources: [Neural Bits]() Topics: [Ollama](), [Local AI](), [Pydantic](), [App](), [Large Language Model](), [Python](), [Code](), [Development]() Tags: [ai](), [code](), [dev](), [examples](), [guide](), [introduction](), [local](), [models](), [ollama](), [practical](), [tools]() ### AI overview A practical tutorial on building a local AI task manager with Ollama and PydanticAI. The application uses local language models and typed agent tools to handle tasks such as adding work, marking tasks complete, and listing overdue items. ### Source excerpt A practical introduction to agent-based architectures using typed models, tools, and runtime context in Pydantic AI. ## Key takeaways from the PyAI conference on AI evaluation, software design, and open-source maintenance DevFeed: [Key takeaways from the PyAI conference on AI evaluation, software design, and open-source maintenance]() Original publisher: [Read original article]() Author: Pamela Fox (noreply@blogger.com) Published: 2026-03-12T06:40:00Z Content type: article Language: en Sources: [Pamela Fox]() Topics: [Artificial Intelligence](), [Python](), [AI-assisted coding](), [MCP](), [SDKs](), [Pydantic](), [FastAPI]() Tags: [agents](), [ai](), [ai-evals](), [code](), [coding](), [coding-agents](), [conference](), [data-science](), [fastapi](), [github](), [maintainers](), [mcp](), [open-source](), [python](), [sdk]() ### AI overview The article summarizes lessons from PyAI conference sessions on evaluating AI systems, designing Python software for maintainability by coding agents, and handling AI-generated pull requests in open-source projects. It recommends validating LLM judges with labeled data and conventional evaluation metrics, using clearer software abstractions, and developing systems to triage low-quality contributions. ### Source excerpt I recently spoke at the PyAI conference, put on by the good folks at Prefect and Pydantic, and I learnt so much from the talks I attended. Here are my top takeaways from the sessions that I watched: AI Evals Pitfalls Hamel Husain 📺 Watch the video recording | 📊 View slides Hamel cautioned against blindly using automated evaluation frameworks and built-in evaluators (like helpfulness and coherence). Instead, we should adopt a data science approach to evaluation: explore the data, discover what's actually breaking, identify the most important metric, and iterate as new data comes in. We shouldn't just trust an LLM-as-a-judge to be given accurate scores. Instead, we should validate it like we would validate a ML classifier- with labeled data, train/dev/test splits, and precision/recall metrics. LLM-judges should always give pass/fail results, instead of 1-5 scores, so that there's no ambiguity in their judgment. When generating synthetic data, first come up with dimensions (such as persona), generate combinations based off dimensions, and convert those into realistic queries. Hamel created evals-skills, a collection of skills for coding agents that can be run against evaluation pipelines to find issues like poorly designed LLM-judges. Build Reasonable Software Jeremiah Lowin (FastMCP/Prefect) 📺 Watch the video recording Write your Python programs in a way that coding agents can reason about them, so that they can more easily maintain and build them. For example, FastMCP v2 SDK was not well designed (bad abstractions) so a new CodeMod feature required 4,000 lines of code. In the new FastMCP v3 SDK (same functional API, different abstractions backing it), the same feature only required 500 lines of code. To make Python FastMCP servers more Pythonic, Jeremiah is developing a new package for MCP apps which includes the most common UIs (forms/tables/charts), called PreFab: https://github.com/PrefectHQ/prefab Panel: Open Source in the Age of AI Guido van Rossum (CPython), Sa ## Validate Environment Variables with Pydantic DevFeed: [Validate Environment Variables with Pydantic]() Original publisher: [Read original article]() Author: Rick Donato Published: 2025-12-02T08:00:19Z Content type: tutorial Language: en Sources: [Packet Coders - Learn Network Automation]() Topics: [Pydantic](), [Environment Variables](), [.env](), [Python]() Tags: [env-file-security](), [environment-variables](), [python](), [settings](), [tips]() ### AI overview A brief tutorial tip showing how to validate environment variables from .env files with Pydantic before they reach an application. It also notes that this can remove the need to manually call python-dotenv. ### Source excerpt Tip! Validate your env vars from your .env files before they hit your application by using Pydantic. This also eliminates the need to call python-dotenv manually. Here's an example: # uv add pydantic pydantic-settings python-dotenv from pydantic import Field, AnyHttpUrl, ValidationError from pydantic_settings import BaseSettings, ## Recent Guides on AI Inference and FastAPI Backend Engineering DevFeed: [Recent Guides on AI Inference and FastAPI Backend Engineering]() Original publisher: [Read original article]() Author: Alex Razvant Published: 2025-11-29T14:02:54Z Content type: article Language: en Sources: [Neural Bits]() Topics: [AI Development](), [FastAPI](), [Inference Performance](), [backends](), [model-deployment](), [Pydantic](), [Retrieval Augmented Generation (RAG)]() Tags: [ai-engineer](), [ai-inference](), [deploy](), [fastapi](), [guides](), [inference](), [production]() ### AI overview A curated collection of recent guides for AI engineers. It highlights FastAPI and Pydantic practices for building maintainable AI and ML backends, along with inference engines and serving frameworks for deploying models in production across different scales and infrastructure. ### Source excerpt A curated list of the most actionable guides I've published in the past months. ## Moving from Django DRF to Ninja API / Pydantic DevFeed: [Moving from Django DRF to Ninja API / Pydantic]() Original publisher: [Read original article]() Author: Michael Sidharta Published: 2025-11-10T05:35:43Z Content type: comparison Language: en Sources: [Kogan.com]() Topics: [Django](), [Pydantic](), [API](), [Python](), [Development](), [OpenAPI Specification](), [Swagger]() Tags: [api-documentation](), [data-validation](), [django](), [fastapi](), [python](), [swagger]() ### AI overview This article examines moving from Django REST Framework API patterns to Django Ninja API and Pydantic. It describes motivations including reducing boilerplate, improving performance for some use cases, using modern Python type hints and data validation, and enhancing developer experience. It also outlines Django Ninja's type-based validation, automatic OpenAPI documentation, performance focus, and simplified endpoint definitions. ### Source excerpt As our project grows, we're always looking for ways to streamline development, improve performance, and enhance the developer experience. Recently, we've been exploring a shift from our traditional Django REST Framework (DRF) API patterns to a combination of Django Ninja API and Pydantic. This blog post will delve into our motivations for this change, the benefits we've observed, and some considerations for others contemplating a similar transition. Why Consider a Change from Django DRF? Django REST Framework has been a robust and widely adopted solution for building APIs with Django. It provides a comprehensive set of tools, including serializers, viewsets, and excellent browser-based API interfaces. However, as our needs evolved, we identified areas where a different approach could offer advantages: Boilerplate Code: While DRF offers powerful abstractions, creating serializers, views, and viewsets can sometimes lead to a significant amount of boilerplate code, especially for simpler APIs. Performance: For certain use cases, the overhead of DRF's serializer validation and rendering can impact performance, particularly in high-throughput scenarios. Modern Python Features: We were keen to leverage modern Python features like type hints and data validation more extensively, which are core to Pydantic. Developer Experience: A more concise and explicit way to define API endpoints and data structures could improve developer productivity and reduce potential errors. Introducing Django Ninja API and PydanticDjango Ninja API Django Ninja is a web framework for building APIs with Django and Python 3.6+ type hints. It's heavily inspired by FastAPI and offers a number of compelling features: Type Hinting for API Endpoints: You define your request and response models using Pydantic, and Ninja automatically validates and serializes the data based on these type hints. Automatic OpenAPI (Swagger) Documentation: Just like FastAPI, Ninja generates interactive API documentation out o ## Build durable AI agents with Pydantic AI and Temporal DevFeed: [Build durable AI agents with Pydantic AI and Temporal]() Original publisher: [Read original article]() Author: David Montague Published: 2025-11-06T00:00:00Z Content type: tutorial Language: en Sources: [Temporal Blog]() Topics: [Pydantic](), [AI Development](), [reliability](), [OpenTelemetry]() Tags: [ai](), [ai-agents](), [how-to](), [instrumentation](), [python](), [reliability](), [type-safety]() ### AI overview This article explains how Pydantic AI combines type-safe agent development with Temporal's Durable Execution. The integration is presented as a way to help AI agents recover from API failures, exceptions, restarts, and deployments while preserving progress and state during long-running or human-in-the-loop workflows. ### Source excerpt Find out how you can combine Pydantic AI's type safety with Temporal's Durable Execution to build prod-ready AI agents. ## Schema-Guided Reasoning: как научить языковые модели последовательно рассуждать DevFeed: [Schema-Guided Reasoning: как научить языковые модели последовательно рассуждать]() Original publisher: [Read original article]() Author: redmadrobot (red\_mad\_robot) Published: 2025-10-31T16:31:44Z Content type: tutorial Language: ru Sources: [Redmadrobot EN](), [Redmadrobot RU]() Topics: [Large Language Model](), [Structured-data](), [Artificial Intelligence](), [JSON](), [Pydantic]() Tags: [function-calling](), [json](), [llm](), [reasoning](), [schema](), [scheme](), [sgr](), [structured-output](), [tag-1cd610c0e518](), [tag-5335b6f99fba](), [tag-b0a411324cb6]() ### AI overview This tutorial explains Schema-Guided Reasoning (SGR), an approach that guides large language models through predefined schemas. It describes how structured output, Pydantic models, JSON schemas, and constrained decoding can make model responses more consistent, transparent, and easier to test, while noting that SGR is not a universal solution. ### Source excerpt LLM умеют многое: генерировать тексты, анализировать документы, писать код. Но на практике их работа часто непредсказуема -- сегодня модель даёт точный ответ, а завтра на тех же данных ошибается, пропускает ключевые шаги или придумывает факты. Для AI-инженеров это системная проблема. Возьмём автоматизацию документооборота: нужно классифицировать договоры, извлекать реквизиты, проверять стандарты. Но модель работает как лотерея -- результат не поддаётся логике или меняется при повторном запуске с одинаковыми данными. Как встроить такой результат в бизнес-процесс? Для решения этой задачи появился подход Schema-Guided Reasoning (SGR). Его активно продвигает Ринат Абдуллин в материалах по работе с LLM. Идея проста и эффективна: заставить модель мыслить не хаотично, а внутри заданной схемы. Это не панацея, но SGR серьёзно снижает количество ошибок, делает процесс прозрачнее, а также позволяет тестировать отдельные компоненты рассуждений. Читать далее ## Durable Digest: September 2025 DevFeed: [Durable Digest: September 2025]() Original publisher: [Read original article]() Author: Temporal Technologies Published: 2025-09-30T00:00:00Z Content type: news Language: en Sources: [Temporal Blog]() Topics: [releases](), [Pydantic](), [Kubernetes](), [Artificial Intelligence](), [Terraform](), [Retool]() Tags: [2025](), [ai](), [announcements](), [aws](), [blog](), [developer](), [kubernetes](), [releases]() ### AI overview Temporal's September 2025 Durable Digest covers recently shipped and preview features, including Pydantic AI integration, task queue fairness, a Kubernetes Worker Controller, namespace tags, and AWS PrivateLink access. It also highlights Retool and OpenPhone projects using Temporal for AI and voice agents. ### Source excerpt Explore Temporal's latest developer updates: questions from the field, new releases, builder spotlights, learning resources, and ways to connect with our team. ## Filter the tools from MCP servers DevFeed: [Filter the tools from MCP servers]() Original publisher: [Read original article]() Author: Pamela Fox (noreply@blogger.com) Published: 2025-09-18T17:35:00Z Content type: tutorial Language: en Sources: [Pamela Fox]() Topics: [Model Context Protocol (MCP)](), [GitHub Copilot](), [agentic-coding](), [vs-code](), [LangChain](), [Pydantic]() Tags: [agentic-coding](), [github-copilot](), [langchain](), [mcp](), [mcp-server](), [openai](), [python](), [vs-code]() ### AI overview A tutorial on filtering tools exposed by MCP servers to reduce LLM confusion, token usage, latency, context-window pressure, and unintended destructive actions. It covers GitHub Copilot in VS Code, LangChain v1, and Pydantic AI. ### Source excerpt What I like about MCP servers: they give me lots of great tools that can make my agents more powerful, with very little work on my side. 🎉 What I don't like about MCP servers: they give me TOO many tools! I usually only need a handful of tools for a task, but a server can expose dozens. 😿 The problems with too many tools: LLM confusion. The LLM will be presented with the tool definition for every single tool in the server, and it needs to decide which tool (if any) is the best for the job. That's a hard decision for an LLM - it's always better to make it easier for the LLM by narrowing the tool list. Increased tokens. The tool call definitions require more tokens, which can cost more money, increase latency, and potentially even go over the context window limit of the model. Destructive actions. A server may include tools that are read-only, just sending down data to serve as context, but many servers expose tools that do write operations, like the GitHub MCP server's tools for creating issues, closing issues, pushing branches, and many more. It's possible your task requires some of those write ops, but you generally want to be very explicit about whether an agent is allowed to take action that can actually change something about your accounts and environments. Otherwise, you can be in for a nasty surprise when the agent took actions that you weren't expecting. (Ask me how I know...) Fortunately, there is almost always a way to configure agents to only allow a subset of the tools from an MCP server. In this blog post, I'll share ways to filter tools in my favorite agentic coder, GitHub Copilot in VS Code, plus two popular AI agent frameworks, Langchain v1 and Pydantic AI. Agentic coding with GitHub Copilot in VS Code Global configuration When you are using agent mode in VS Code, configure the tools by selecting the gear icon near the chat input window. That will pop-up a window showing all your available tools, coming from both installed MCP servers and VS Code exte ## Temporal and the next frontier: Scaling AI reliably DevFeed: [Temporal and the next frontier: Scaling AI reliably]() Original publisher: [Read original article]() Author: Samar Abbas Published: 2025-09-03T00:00:00Z Content type: article Language: en Sources: [Temporal Blog]() Topics: [systems](), [AI Infrastructure](), [agentic workflows](), [SDKs](), [OpenAI](), [MongoDB](), [Pydantic](), [Retrieval Augmented Generation (RAG)]() Tags: [agentic-workflows](), [ai](), [ai-infrastructure](), [announcements](), [context-engineering](), [developer-community](), [mongodb](), [openai](), [rag](), [sdk]() ### AI overview Temporal describes how its durable workflow infrastructure is being applied to production AI systems, including agentic workflows, multiple AI models, stateful processes, and integrations with OpenAI, MongoDB, and Pydantic. ### Source excerpt At Temporal, we've always focused on one thing: making it simpler to build systems that work and keep working -- reliably at scale and in the messy reality of prod. That mission matters more today than ever. ## Building Agents With Heroku AI and Pydantic AI DevFeed: [Building Agents With Heroku AI and Pydantic AI]() Original publisher: [Read original article]() Author: Anush DSouza Published: 2025-08-08T16:33:37Z Content type: tutorial Language: en Sources: [Heroku]() Topics: [Heroku](), [Pydantic](), [AI Agent](), [Agent Framework](), [Model Context Protocol](), [A2A protocol](), [Python]() Tags: [a2a](), [agents](), [ai](), [ai-agent](), [claude](), [fastapi](), [heroku](), [heroku-ai](), [managed-inference-and-agents](), [mcp-on-heroku](), [model-context-protocol](), [model-context-protocol-mcp](), [news](), [product-features](), [python]() ### AI overview This tutorial explains how to build production-grade AI agents with Heroku Managed Inference and Agents and the Pydantic AI Python framework. It covers Claude model integration, secure code execution, OpenAI-compatible APIs, environment-variable configuration, MCP client and server support, and exposing agents through the A2A protocol. ### Source excerpt Building production-grade AI applications can be complex, but with Heroku and Pydantic AI, developers gain a powerful and reliable solution for integrating advanced AI capabilities. Heroku makes it easy to integrate AI into your applications with Heroku Managed Inference and Agents. With a single click, you can attach powerful Large Language Models like Anthropic's Claude [...] The post Building Agents With Heroku AI and Pydantic AI appeared first on Heroku. ## Structured Output with LangChain and Llamafile DevFeed: [Structured Output with LangChain and Llamafile]() Original publisher: [Read original article]() Author: brakmic Published: 2025-06-22T16:50:12Z Content type: tutorial Language: en Sources: [Harris Brakmic - Coding]() Topics: [LangChain](), [llamafile](), [JSON](), [llama.cpp](), [Parser](), [Pydantic]() Tags: [ai](), [code](), [coding](), [cpp](), [how-to](), [import](), [json](), [langchain](), [llama-cpp](), [llamafile](), [llm](), [local](), [local-llm](), [models](), [properties](), [python](), [structured-output]() ### AI overview A tutorial on using LangChain with Llamafile to produce structured JSON output from a local large language model. It uses JsonOutputParser, PromptTemplate, and a Pydantic model because Llamafile does not provide a with_structured_output method. ### Source excerpt Learn how to extend Llamafile with LangChain's JsonParser to produce clean, structured JSON output