What are you building and whats your architecture like?
retained text
I'm curious what people here are building beyond basic chatbots/agent loops, and what your architecture looks like. I'm building Dewey, basically an AI-assisted research/knowledge-work system for a one-person research business. Right now it's less one giant agent and more a collection of systems around a strong generalist (mostly Codex): persistent project state, research/evidence workflows, provenance, reusable methods, deterministic checks, recovery/repair, and governance that helps control what Codex is working on and what it's authorized to do. A big principle I'm exploring is progressive determinism: let flexible models handle messy work, observe what repeatedly works, then turn stable pieces into normal software, validators, or tightly bounded workflows. The flexible version stays available for cases outside the automated one's niche. I'm also trying to separate the work from the worker. Ideally the same bounded operation could be handled by Codex, a cheaper model, a specialist agent, a script, or a human without the project's state and rules living inside that worker's context. The rough direction is: persistent state → bounded task/context → appropriate worker → structured result → validation → state update → next work The biggest metric I care about isn't token cost; it's human attention per unit of valuable output. Still very much evolving, and I'm currently mapping the existing system before deciding what the next architecture should actually be. What are you building, and what architecture have you ended up with? Particularly interested in what you initially thought needed agents but eventually turned into normal software.
document fetched 2026-10-04 00:18 · canonical https://reddit.com/r/AI_Agents/comments/1wwkjyc/what_are_you_building_and_whats_your_architecture