Loops vs. graphs: a new way to structure agent workflows around company memory
femke_plantinga · x · 2026-07-23
The post lays out two ways to run an agent on the same support task: loops and graphs.
- Loops: you define the goal and quality bar, and the agent owns the path, drafting, checking, and retrying until it passes.
- Graphs: you define the workflow up front — read → triage → draft → QA → send — and the agent fills each node.
Both rely on a shared company brain: past tickets, policies, and product docs that each step can pull from as the work moves through the process. The takeaway is simple: use loops for one-off, exploratory work; use graphs when the work is repeatable and the path is already known.
Related event: Evolution of AI Agent Architecture: From Loops to Graph Engineering(8 posts)→
More from coding & agent
- Team-level AI agents: where should shared context and history live? — Al_Grigor · 2026-09-11
- Chaining dependent MCP tool calls: no rollback, duplicate risk — agentrsdg · 2026-09-11
- DeepMind-led paper makes design docs the source of truth, code disposable — SMART regenerates in 1.5-3h for ~$100 — Roger_M_Taylor · 2026-09-11
- Agent-built classifier labels 192k docs for $0.70 vs $13-26 with frontier LLMs — vanstriendaniel · 2026-09-11
- MathModelAgent gains traction: auto-solves math modeling and writes a submission-ready paper — jihe520 · 2026-09-11
- alphaXiv open-sources OpenResearch to run parallel research agents with any model — alphaXiv · 2026-09-11