Posted inBlog
Rules and Hooks in AI Coding Agents
Often when you start putting an AI coding assistant to work, the configuration simply means writing instructions. You put them in a markdown file — `CLAUDE.md`, `agents.md`, or whatever the…
Posted inBlog
Combine Compute, Don’t Choose It: What Multi-Model Harnesses Make Possible
For a long time, the standard advice when adopting AI tooling was straightforward: pick the best model for your use case and use it consistently. That made sense when releases…
Posted inBlog
The Three-Part Job Behind AI’s Widest Salary Bands
The AI job market has been generating new titles faster than hiring managers can agree on what those titles mean. Forward-deployed engineer is a good example. Salary bands for the…
Posted inBlog
The Slop Apocalypse and Why Spec-First Still Wins
When agentic coding first landed in engineering teams, the early promise was simple: offload the tedious parts and keep the interesting ones. What nobody anticipated was that the tedious parts,…
Posted inBlog
You Bought the Agent to Get Time Back. Here Is Why Your Calendar Filled Up Instead.
When the first wave of AI assistants arrived, the mental model was straightforward: delegate a task, get time back. Agents were supposed to be the escalation of that promise —…
Posted inBlog
From Copilot to Fully Agentic SDLC
Most engineering teams have been adding AI tools at the edges of their workflow — a coding assistant here, a code review bot there. Uber's engineering team has been building…
Posted inBlog
Your Agent Evolved. Did Your Evals?
For a long time, the question asked about an AI system was: "Does it give the right answer?" That question makes sense when a system is a single model call…
Posted inBlog
Start With a Workflow, Then the Agent
The idea of agents "automating our jobs" has gone around for a while, but most examples stay abstract. Niels Rogge at Hugging Face followed a concrete, reproducible, and honest approach…
Posted inBlog
Agent Infrastructure Is Solved. Now Comes the Context Problem.
The Infrastructure Tax Two years ago, building a production-grade required infrastructure: state persistence, sandbox isolation, observability across disconnected systems. A small team could dedicate 3 months just on that. Such…
Posted inBlog
Safe Autonomy Is an Environment Problem, Not a Model Problem
When AI agents first appeared in production workflows, the permission question felt manageable: give the agent read access to a few APIs, make sure it could not write anything it…