Organizations using Oryggi Access Control Systems had powerful databases tracking every employee, door access, visitor entry, and attendance record — but extracting insights required writing raw SQL. Only IT staff could answer basic questions like “How many employees joined last month in IT?”

Traditional template-based query tools achieved just 33% accuracy. Managers needed reports that took days. HR needed to revoke door access but had to file tickets. The data was there — but it was inaccessible to the people who needed it most.

On top of that, India's new DPDP Act 2023 meant every data query now required proper consent tracking. There was no compliance infrastructure in place.

I designed AiXpert as a multi-tenant SaaS platformwhere any employee can type a question in plain English and get an accurate answer from their organization's database — without knowing SQL, without waiting for IT, without risk of accessing another tenant's data.

The AI pipeline uses Google Gemini 2.0 Flashwith a RAG layer (ChromaDB) that understands each tenant's unique schema. Few-shot learning means the system improves with every query. The result: 90%+ accuracy vs 33% from the previous template approach.