A solo founder runs 16 AI agents to sell B2B software, and code beats prompts on guardrails
mastafied · reddit · 2026-07-22
A solo founder running industrial-insulation software alongside a full-time job says he now orchestrates roughly 16 AI agents through Paperclip, with Claude and Codex doing lead research, enrichment, cold email, CRM updates, SEO posts, blog publishing, LinkedIn comments, Reddit scouting, call prep, and call transcription.
What worked
- 214 cold emails produced a 3.3% reply rate, but the pipeline still uncovered and qualified the first pilot customer end-to-end.
- A lead clicked but did not reply, so he called instead; that became the best conversation in the campaign.
What broke
- The copywriter agent hallucinated product capabilities.
- Prompt instructions helped but did not eliminate the problem, so he added a QA agent that checks every draft against a hard allowlist of claims.
- Non-negotiables moved from prompts into code: the send endpoint now blocks do-not-contact addresses at the API layer.
Interesting failure mode
- The agents even diagnosed flaws in their own instructions, explained why templates were causing rejected drafts, and proposed fixes that he reviewed as diffs.
He ends by asking how others build stronger hallucination gates for customer-facing agent output, and whether structured claim extraction against a product fact base is better than “one LLM judging another.”
More from coding & agent
- Alex Townsend posts 200 open problems in numerical linear algebra for humans and AI agents — IgorCarron · 2026-09-11
- Kimi K2.8 Preview rolls out: near-K3 coding performance, 1M context for all tiers — teortaxesTex · 2026-09-11
- Looking for a classifier of software engineering task shapes to pick models per task — StewartalsopIII · 2026-09-11
- Steal this idea: prompt-to-hardware where agents assemble custom devices — paraschopra · 2026-09-11
- Model Is the Least Interesting Part: A Guide to Six Core AI Architectures from RAG to Multi-Agent — goyalshaliniuk · 2026-09-11
- Non-coder builds layered memory architecture: 20k tokens tracks a year of agent conversations — matteoianni · 2026-09-11