What I'm Working On

Building

ISNAD v2.23.0 — chain-scoped content grounding, content-madār fingerprint tightening, and a warm default registry shipped. Now running the preregistered model-drift leaderboard: hallucination originates at memory-generation (hop 1), not chain depth. LangChain + MCP integrations live; paper on arXiv (2607.24117).

At SHCV.IT

AI System Architect — co-architecting a self-maintaining enterprise knowledge platform. Designing multi-agent AI systems, knowledge synthesis pipelines, and production AI architecture for the DACH region.

At Islam & AI

Founder — shipping native iOS & Android apps plus the backend on top of the live assistant: Qur'an + 600k hadith RAG, memory, and deep research. The trust problems here, combined with my multi-agent systems work, became ISNAD.

At Entropic

Founder — pivoted into an AI Operating Brain company for founder-led agencies. Installing intelligence layers above existing tools that surface what's at risk, what's due, and what needs attention.

Reading & Learning

Provenance and trust literature — the intellectual lineage behind ISNAD. Jøsang subjective logic, truth discovery, hadith transmission criticism. Deepening the framework's theoretical foundations.

Global Tech Leverage

Growing the private network of engineers moving into global roles. 30+ engineers. Applications open.

Open To

Selective full-time roles if the right opportunity comes, plus fractional CTO engagements, research advisory, and technical co-founder conversations. Fully committed to SHCV, Islam & AI, and Entropic day-to-day.