AI Systems

AI systems I've implemented

Every system here started from a real pain in a real department. I sat with the team, understood the process deeply, and built a solution that works in the field today.

Customer-complaint agent for the QA department

A system I built end to end, through months of tuning and feedback with the QA manager: an agent that drafts a complaint report at the level of the manual one, one to one. It knows the product catalog and SKUs, follows the quality procedures, and cross-references recurring complaints to catch patterns in time.

ClaudeQA ProceduresRegulated Industry

AI workshop for management, training for teams

Each department head mapped recurring tasks from their own work, and I built the content around them: smart Excel and formula workflows in Claude, and dedicated project folders grounded in real business data. Since then, follow-up training for sales, product, finance, procurement and engineering teams, with structured projects and prompts per team.

AI workshop for Sion Medical management
ChatGPTGeminiNotebookLMClaude

Real-time sales agent

Field reps were losing time on every stock or order query. I mapped the daily questions with them and built a chat agent that answers in seconds, straight from the ERP: inventory, open orders, quotes and customer data.

Priority ERPAPI

Delivery-ETA automation

Procurement was keying delivery forecasts by hand from Excels and emails. I built a system that listens to supplier emails and updates expected stock arrivals in Priority automatically. It cuts about a third of a role's manual data entry.

Email AutomationPriority

AI Agents course

A hands-on course on building agents that run tasks end to end, led by Tom Even of agents&me. I built a personal agent connected to my email and systems, generating reports automatically. It's been part of my routine since.

Let's talk