Building an AI-Powered Operating System for an Innovation Agency (Knowledge OS + Scouting Agent + Augmented PM)

I designed and deployed a full AI-driven Operating System for an innovation-focused agency managing complex, multi-country projects (executive programs, expert coordination, Learning Expeditions, etc.).

The system was built around three main pillars:

1) Knowledge OS – Capture → Store → Retrieve

Integrated automated meeting capture (Fireflies or equivalent).

Built a complete Notion knowledge hub (meetings, experts, clients, projects).

Automated AI-generated summaries with thematic tagging.

Implemented a RAG-based semantic search system to query internal history, e.g.:

“What recurring logistical issues did we face in our recent Japan programs?”

Delivered documentation, SOPs, and team training.

2) Scouting Agent – Automated Research & Enrichment

Built an end-to-end scouting workflow using Browserbase Director / alternatives.

Automatic enrichment of startups/experts using APIs (Clay / LinkedIn / Crunchbase).

Structured Airtable CRM for experts, companies, and contacts.

Automatic generation of AI briefings for the team.

Added quality scoring, monitoring, and error-handling systems.

3) Augmented Project Management (PM) System

Evaluated multiple PM tools: ClickUp, Wrike, Asana, Miro.

Created complete LEX project templates (phases, dependencies, milestones).

Implemented AI automations: alerts, risk scoring, reminders, weekly reports.

Synced Notion ↔ PM tool ↔ Airtable ↔ Make for seamless operations.

Connected Miro workshops to task creation and backlog generation.

Result:

A fully operational internal OS where knowledge is centralized, scouting is automated, and project management becomes proactive instead of manual.