# ai-agents-in-action Published articles for ai-agents-in-action. This is one page of public article previews, not the complete archive. Follow Next page to continue. Summaries are not the original full articles. ## How Instacart Built Blueberry, a Slack-Native On-Call Reasoning Harness DevFeed: [How Instacart Built Blueberry, a Slack-Native On-Call Reasoning Harness]() Original publisher: [Read original article]() Author: Karthik Halukurike Published: 2026-07-14T16:39:20Z Content type: tutorial Language: en Sources: [Instacart]() Topics: [incident](), [Slack](), [context](), [dashboards]() Tags: [ai-agent](), [ai-agents-in-action](), [context](), [incident-response](), [on-call](), [reasoning](), [slack](), [software-development]() ### AI overview Instacart describes Blueberry, a Slack-native on-call reasoning harness that auto-triages qualifying alerts and supports follow-up investigations. It is designed to help teams reach a shared understanding and test theories more quickly during on-call work. ### Source excerpt How we built a Slack-native on-call reasoning harness at Instacart that shortens time to first insight, speeds up theory testing, and turns tribal knowledge into reusable infrastructure. Key Contributors: Karthik Halukurike, Gabe de Oliveira, Hassan Jallad, Alan Wong On-call work is a race to turn noisy signals into shared judgment. The hardest minutes of that race aren't the ones spent fixing the bug -- they're the ones spent figuring out what the bug even is, while everyone in the thread is asking the same question from a slightly different angle. Blueberry is the system we built at Instacart for those several minutes. It lives in the Slack thread where the team is already coordinating, picks up each alert as it fires, and lands a grounded explanation back in the thread in about three minutes -- thousands of times a month. On-call is mostly a clarity and speed problem Most on-call pain does not start with a major incident. It starts in the noisy window right after a page fires, when the thread fills with links, dashboards, guesses, and partial context. Someone asks whether a deployment caused it. Someone else asks how broad it is. A third engineer joins and asks what is going on. In that moment, the hardest problem is often not deep root-cause analysis. It is getting to a shared understanding quickly enough that the team can make a good next move. For us, that made two metrics matter most: time to first insight (TTFI) and time to test theories (TTTT). TTFI comes from auto-triaging every qualifying alert the moment it lands. TTTT comes from engineer-initiated on-demand investigations and follow-ups as the investigation deepens. Blueberry is a Slack-native on-call reasoning harness Blueberry was built for that first window of confusion. Blueberry lives in Slack because that is where on-call engineers already coordinate, ask clarifying questions, and make decisions under pressure. We wanted the system to work inside the shared operational conversation so the full team ## How We Built an AI Agent to Clean Up Dead Code After A/B Tests DevFeed: [How We Built an AI Agent to Clean Up Dead Code After A/B Tests]() Original publisher: [Read original article]() Author: Aseem Upadhyay Published: 2026-06-23T10:54:43Z Content type: article Language: en Sources: [Housing.com]() Topics: [AI Agent](), [experiments](), [Code](), [Pull Request](), [context](), [Android]() Tags: [agent](), [ai](), [ai-agent](), [ai-agents-in-action](), [automated](), [code](), [concurrent](), [experiment](), [experiments](), [infrastructure](), [jira](), [llm]() ### AI overview Housing.com describes building an AI-agent pipeline to help clean up code after A/B tests. The workflow interprets experiment tickets, reports experiment status, and applies instructions to code, while using scripts for deterministic steps and an LLM where judgment is required. ### Source excerpt Photo by Microsoft Copilot on Unsplash At Housing.com, running product experiments is a continuous cycle. A/B tests go live, collect data, and eventually reach a conclusion. That's the exciting part. Then comes the mundane reality where someone has to clean up the code .i.e. remove a feature flag, promote a winning variant or revert the loser variant and finally raise a change request. Sounds simple? Maybe Is it tedious and quietly expensive? Yes! lifecycle of a taskThe Problem Worth Solving An experiment conclusion ticket typically lands on an engineer's desk looking something like this: Experiment: show_listing_map_widget Platform: Android Result: Negative - revert to control The job of the assigned engineer involves four distinct steps: Find every reference to the flag across the codebase. Delete the losing variant's code path. Trace every side-effect that only existed to support that variant Commit, open a PR, and comment on the Jira ticket. Step 3 is where the trap lies. Be it applying or removing a change, changing all the infrastructure code dependent on it could increase the complexity and risk of creating technical debt. But what if we automated a part of it? The AI Agent Pipelineupdated AI enabled lifecycle The problem statement became simple: Let stakeholders own the trigger. We built two agents to make it happen, 1. to interpret tickets and report experiment status 2. to take the instructions and code. Then came the hard part. Navigating Roadblocks The real complexity lies in building an AI agent that runs autonomously and serves different users across different use cases We found ourselves wrestling with questions we hadn't fully anticipated: How do we optimise on the tokens used per request? How do we handle concurrent requests? How do we ensure that the consistency in the output? So we went looking for answers.. Optimising Tokens per request Not every step needs AI. At each point in the workflow, we asked one question: is this operation deterministic