I ran my job search as an operated pipeline. Then I open-sourced it.
The story of career-ops: 740 evaluated, 68 applications, 12 interviews, 1 offer. Open-sourced. And the CEO found me through the system, not a résumé.
Case studies on AI agents, multi-agent systems, LLMOps and open source — by Santiago Fernández de Valderrama (santifer).
The story of career-ops: 740 evaluated, 68 applications, 12 interviews, 1 offer. Open-sourced. And the CEO found me through the system, not a résumé.
A fleet of Claude Code agents maintains career-ops: triage, tests, review briefs and releases on evenings and weekends. The system, one documented day, and the playbook.
Case study: AI job search tool built as a multi-agent system. AI resume builder, multi-dimensional scoring, automated applications with HITL. 740 offers evaluated.
Case study: production LLMOps with agentic observability, 6-layer defense, 71 evals, voice mode, and a closed-loop that generates tests from real failures.
Case study: 4,730 static landing pages from Airtable as headless CMS with DataForSEO crawl budget optimization and Astro SSG. 2M+ impressions, 19K+ clicks.
Case study: custom Business OS with 12 Airtable bases, 2100 fields, and n8n saving 170h/month at a phone repair business.
Case study: omnichannel AI agent with sub-agents, tool calling & multi-model orchestration (n8n + ElevenLabs). 90% self-service.
n8n cheat sheet for Product Managers: automate sprint reports and classify feedback with AI. 2 free importable workflow templates. Step-by-step tutorial.