Erwann Bestard

At Dashdoc, I build AI-powered internal tools and agents to automate operations and make the team more effective, from idea to production (spec → prototype → rollout → monitoring). AI products & agents (last 6 months) • Built an internal RAG chat assistant on top of our Notion and internal documentation to help teams answer product, process, and technical questions in seconds instead of searching manually. • Developed an AI agent to clean and reformat CSV exports (column mapping, normalization, validations) to dramatically speed up customer onboarding and data imports. • Designed and shipped a multi-agent AI system able to perform various operations in our core product (data checks, automation, and bulk actions), with safe-guard rails and human-in-the-loop validation. Platform & backend • Set up and maintained CI/CD pipelines, improved deployment flows, and strengthened production monitoring (logs, alerts, dashboards). • Participated in incident response during red alerts, helping restore service quickly and improve our post-mortem processes. • Contributed to backend APIs and business logic when needed, mostly around data flows and integrations. Skills: AI Agents, LLMs, RAG, Python / TypeScript, Postgres, CI/CD, Monitoring, Incident Response

Implemented an ETL Solution: • Conducted a benchmark of various technologies (DuckDB, Clickhouse, BigQuery) • Built a proof of concept using Clickhouse with PeerDB • Ultimately deployed materialized views in Postgres with periodic refresh, following a star schema design

Mobile and web development with React.js and React Native, TypeScript