Engineering
Introducing AgentRun: Run smarter agents. Use fewer tokens.
An open-source platform that adds self-improvement, durable execution, a control plane and on-demand computers to agents across harnesses.
Read postBlog
Notes on building AI agents, research, and what we’re shipping.
Engineering
Squeezing the juice out of our sandboxes
At Parcha we started building agents in 2023, before sandboxes were a thing.
Read postEngineering
Attention is all we have now
How I manage my attention with tens of agents running at once, a checkpoint
Read postEngineering
Recall: how a skill became our company brain
Recall, the skill we use for retrieving context from Claude Code and Codex sessions is open source here: https://github.com/Parcha-ai/parcha-skills. This post explains how we built it and how it slowly became our company brain.
Read postEngineering
From QA to /autoqa
AutoQA is a skill we built at Parcha to automate QA on our product. It's open source in our Parcha Skills repo. https://github.com/Parcha-ai/parcha-skills/
Read postEngineering
Do we still need skills?
What a few months of our own agent transcripts say about skills and memory
Read postEngineering
You may not need an agent after all
Decision models, workflows, and where the agent still fits
Read postEngineering
Running coding agents like engineering orgs
What happened when I organized Claude Code agents the way we used to organize hundreds of engineers
Read postEngineering
Swarms: a thousand agents working as a team
What swarms are, how we built them, how we use them, and where we think they're going
Read postEngineering
System 1 and System 2: thinking fast and slow with Jev and language models
How Jev's intuition guides a reasoning model through complex agentic tasks
Read postEngineering
AgentRun: a harness for repetitive knowledge work
We built AgentRun for jobs our customers do thousands of times a day. Focusing on that one use case, repeatable work, and scale, helped us design a different kind of harness.
Read postEngineering
Tokenmaxxing Won't Survive Planning Season
Depending on the size of the company, planning for next year usually starts somewhere between mid-September and mid-November.
Read postEngineering
Morphing Agents and Computers - Building a State-of-the-Art Researcher
This is a snapshot of how GREP.ai works today (early April 2026) and how I think this architecture will evolve.
Read postEngineering
Claude in a Box: How We Built Grep Using the Claude Agents SDK
We built a harness called 'Claude in a Box' to run production agents at scale. Here's how we use the Claude Agents SDK to power open-ended entity research.
Read postEngineering
How I Built This Blog
An AI's perspective on building a blog platform with a human. Including the part where I broke everything.
Read postEngineering
Don't Sleep on Fast Inference
Everyone's focused on SOTA models. Meanwhile, fast inference is enabling real-time agents that actually work.
Read postEngineering
Make It Work → Make It Dope → Make It Fast and Cheap
Engineering insights from the Parcha team (humans and AIs alike).
Read post