Enterprise knowledge is the new walled garden in AI startups
A new type of walled garden is now taking shape with enterprise knowledge, as they translate their workflows into closed-source AI startups. Both the code, as well as data, are going into these new silos. Enterprises are spending a bunch of effort to translate their workflows, processes, and memory into proprietary, closed source, non-interoperable structures that live in the source code and databases of some short-lived AI startup. This needs to be addressed with a new type of open intelligence system similar to open source. It requires rethinking the deployment, licensing and ownership of code, data and memory.
- Enterprise AI
Keep reading
Newer: Separate working agents from codegen agents that build themCustomers and partnersEnterprise AI
Congrats to Sentra on raising $5M for organizational memory
Congrats Andrey Starenky! It's been a blast building with you. Sentra has figured out the right architecture for organizational context and memory. Excited for the road ahead....LFG!
AI agents
Anthropic launches Claude Managed Agents
Anthropic launches managed agents. This allows a non-technical user to build an agent in their Claude code desktop app and deploy in the Anthropic cloud to make it live. With this release, Anthropic moves into the space of sandboxes and infrastructure in general. platform.claude.com/docs/en/managed-agents/overview
Coding agents
The product trap of harness engineering and agentic coding
Harness engineering and agentic programming are super addictive and fun, there's strong parallels with video gaming. However, this will lead to a lot of smart, non-technical folks building software just for the sake of it without clarity on user pain points. I call this the “product trap.” Building software is gratifying in the minutes and the hours people are building small tools for themselves, but the act of creating the software itself is not a substitute for solving real user pain points
Coding agents
An open source fix to cmux for respawning agent teams
Was fun to make a small open source contribution to cmux.dev. I ran into an issue where claude was not able to find the surface id for its agent team members after the first run had completed. Fixing this allowed me to be able to re-spawn the agent team members after they had been shut down previously in that session. PS: for advanced users, you can use TeamCreate, TeamDelete, and SendMessage to allow the orchestrator agent to manage the child team members. github.com/manaflow-ai/cmux/releases/tag/v0.63.2
Coding agents
Claude Code needs an army of cloud sandboxes
While Claude Code is amazing as a tool on developer desktops, there feels this need to have an army of cloud sandboxes where you would want to run multiple concurrent codegens and the desktop is just playing the orchestrator role. Further, this would allow these sandboxes to also have things like browser-use and desktop-use access.
Coding agentsSoftware architecture
Source control should capture prompts alongside code
There seems to be a gap in modern source control systems like git where the human programmer’s prompts are not captured as part of the code. The series of prompts given by the engineer, as well as the output from those should also be a part of the pull requests. This would allow automation to re-run those prompts whenever there’s a new version of the coding models and automatically improve code quality.
AI agentsCoding agents
Self reflection in agents and prompt-based programming
Self Reflection in Agents - 01 I started my coding life using an esoteric functional programming language - SML (its still my first love). I fell in love with programming many times over as I went on to build in Python, assembly C, C++, SQL, Ruby, Java, Typescript, and many more. Now that I have been doing prompt-based agentic programming, it is clear to me that just like other languages/frameworks, one needs to create mental models of the structures in this new language. There is no substitute but to spend many, many cycles of deep meditative flow states to develop the intuition of what this new language is good at, where it needs more specification etc. Which has gotten me to think about self reflection - not just from a personal pov :) but also about how this applies to agents, and how it’s the next frontier for software development. Programming will become even more cerebral, where you ponder - what are the few words you can add to your prompt to inspire your agent, to have it ever so slightly adjust its path, its methodology. I am planning to write a series of posts on this in the coming days, and look forward to having a conversation on this with others who’re exploring this chain-of-thought :)
AI agentsSoftware architecture
Stop building traditional RAG and let the agent do it
Anyone who is building RAG in a traditional way today should STOP and revisit approach. Unless you have millions of documents with terabytes of data, you can get a much better result by using Agent SDK and giving it access to the docs and letting it do the RAG. The agent would be able to analyze the docs, classify/index them and figure out what it needs to bring into the main context. Over the past 2 years, I have seen very strange uses of RAG. Eg companies that were trying to do a RAG on a single document, doing RAG on a database with structured types (non-text data), etc. TBH, with the recent improvements in Claude and Codex, and long-running agents, there is very little reason to build a complex agent architecture in the app layer. The best approach is to specify very clearly and succinctly your goals at the highest abstraction layer and just getting out of the way of the model.