Written by humans
Blogs, research and in-depth write-ups explored by our team.
Sign up for updates
by Tyler Edwards
Frontier models are overkill for most agent tasks. Specialised small language models beat them on cost, latency, and accuracy.
by Pritam Soni
Learnings from building, deploying and improving agents in production.
by Sam Brunt
Application layer companies are training their own models. Why?
by Rohit Gupta
Every agentic system can create its own training data.
A perspective on when to tinker with the prompt, and when to train your own model.
What's the difference between SFT, LoRA and Distillation?
Own your intelligence, or rent the API?
An agent has a model, loop, tools and memory.
Observing your agent is table stakes.
Your agent is the most documented system in your engineering org and yet you can't answer simple questions.
