kapa.ai

A year ago, we partnered with kapa.ai to launch an AI assistant for ODK’s documentation. Since then, it has answered nearly 24,000 questions from 2,800 people and maintained an 85% helpful rating. One thing I’ve learned over the past year is that AI isn’t just about creating new things. Some of the biggest gains come from helping people find, understand, and apply the resources we’ve spent years creating. If you have a question about ODK, give the Ask AI feature a try: https://docs.getodk.org. You might be surprised by how helpful it is.

Your technical support team keeps solving the same ticket twice. Not because they're careless - because the fix from six months ago lives in a closed ticket nobody will ever find. We just shipped Salesforce Cases as a data source for kapa.ai - the last big ticketing system after Zendesk and Jira Service Management. Why? Your case history is verified problem-and-solution pairs written by your best support engineers. Docs tell you how the product should work. Tickets tell you how it actually breaks, and how to fix it. The most common ways teams use it: 1/ Surface similar cases and past resolutions the moment a new ticket comes in 2/ Give L1 support answers that used to require an escalation 3/ Spot recurring issues that reveal exactly what your docs are missing 4/ Build your own support automations (auto-triage, agent-assist, account summaries) with kapa as the grounded retrieval layer Your support history is a goldmine. Stop treating it like an archive. Docs in the comments 👇

We are launching kapa.ai for Agents: all your product knowledge, in one tool call. Claude Fable shows that context = final bottleneck. Here's why: Say you've built an agent in your SaaS app with some tools. Then a user asks "How do I enable SSO?" but you haven't given the agent a tool to fix that. It doesn't matter how smart your agent is, it hits a dead end. Instead, with Kapa, you can easily add a single knowledge search tool so your agent can read your docs, code, tickets. Real-world agents use this tool in +40% of interactions to improve planning and avoid dead ends. TL;DR: we spent 3 years building the best agentic retrieval platform that: → Finds the right source ~2x more often than web search or a DIY RAG pipeline → Tells you what your agent couldn't answer, and exactly how to close the gap → Connects 30+ sources in one click, synced in real time so knowledge doesn't go stale → Works with any agent: product copilot, support agent, or Claude Code Teams like Port.io, Airbyte, and CircleCI have built in-product copilots, support agents, RFP tools, and coding assistants using Kapa for product knowledge. Get a free API key today at kapa [.] ai / agents

Announcing the BIGGEST kapa.ai launch of 2026: code as a data source. "Wait, how hard could this be?" Trust me... We've been trying to crack code ingestion for over 2 years. We tried so many times. 30+ customers (and 75% of our largest logos) maintain some form of public code, and specifically asked for it. Because combining docs and code in one RAG system is insanely hard. A single repo can have 10x more content than all your documentation combined. And you can't just dump everything in to one giant .MD file. For 2 years, models just weren't good enough. We tried. And tried. And tried again. But in the last few months, something clicked. Our research team saw the opening and absolutely cooked. We finally did it. What we built: code-aware chunking that parses your codebase down to every function, every class, every method definition - directly integrated into our agentic RAG pipeline that understands the structure of your entire codebase. And STILL returns a cited answer with time to first token in under 2.5 seconds. The results are wild. In our experiments we found that 50-80% of your users' questions can be answered from source code alone. Documentation tells you the "why." Code tells you the "what." Now your AI agent speaks both languages. For technical writing teams, this meaningfully changes your roadmap. If your AI can pull implementation details directly from code, your writers can stop documenting every function signature and edge case. They can shift to the high-leverage stuff humans are great at - tutorials, architecture guides, the "why" behind the code. In short, this is going to change a lot, for a lot of teams. We're rolling this out now. Link in comments.

Exciting news! We're sending Kapa's allstars to the World Cup of semiconductor conferences: Embedded World in Nuremberg. If you live and breathe datasheets, come find us. One of my favorite things about these industry-leading events is seeing our industry-leading customers. Silicon Labs, Nordic Semiconductor ASA, and Espressif Systems all power billions of IoT devices. They've got the most complex developer documentation on the planet. They all use kapa.ai to turn that documentation into an AI assistant their users love. We'll see them at Embedded World, and we hope to see you too. If you're one of the 32,000 engineers or 1,200 exhibitors at Embedded World, come find us at Hall 4A / Booth 4A-101. We'll show you how AI can handle even the most technical docs. Looking forward to seeing everyone there!

Ever wonder why Claude Code is so good? It has access to your codebase. It knows your project. It has context. Now every company is trying to build the same thing for their product. Companies like Grafana and Amplitude have built AI sidebars that can do things like create_dashboard(), run_cohort_analysis(), and more. But here's the thing. Just like Claude Code needs your codebase, these copilots need to know your product. And that knowledge? It lives scattered across your docs, API references, help centers, wikis, GitHub repos, and forums. Without it, your copilot can't reason. It just guesses. We've solved it with one simple call. POST /retrieval { "query": "what alert evaluation intervals are supported?" } That returns the most relevant chunks from your entire product knowledge base, ranked, with sources. Behind that one call: state of the art low latency multi step agentic retrieval with query decomposition and reranking. 50+ managed data source types, auto synced, always fresh. Optimized for product knowledge. Hardened across 200+ production deployments over 3 years. You don't see any of that. You just get the right chunks back, fast. Plug it into your copilot as an API call. Or use search_product_knowledge() as a hosted MCP tool call. Either way, your agent goes from generally smart to product expert. You build the copilot. We handle the knowledge layer. Link below to get started.