Mohammad Jahangir

Mohammad Jahangir — Riyadh, Saudi Arabia

AI & ML Development

AI/ML work here is applied, not academic: it's integrated into existing product flows to remove a specific manual step. In Opperiq, that's face/selfie verification (YOLOv8, InsightFace) for secure attendance and OCR for receipt/expense processing. On the MMR data-center program, it's AI-assisted document classification and routing between Aconex and SharePoint, and structured extraction of review comments from technical submittals.

What this looks like in practice

  • Building an AI-powered face/selfie verification service (YOLOv8, InsightFace) with model caching for low-latency mobile verification
  • Building OCR-based receipt processing that extracts expense data directly into a finance workflow
  • Building AI-assisted document classification and routing between Aconex and SharePoint for a construction program
  • Training and deploying ML models with end-to-end pipelines (preprocessing, feature engineering, model serving via Flask and FastAPI)

Frequently asked

Does Mohammad Jahangir build AI/ML features or just call third-party APIs?

Both, depending on the problem — Opperiq's face verification uses YOLOv8 and InsightFace directly with model caching for performance, while other workflows use trained models served through Flask/FastAPI pipelines.

What's a concrete AI feature currently in production?

Opperiq's face/selfie verification for employee attendance, and OCR-based receipt processing that feeds directly into the platform's finance module.

Technologies

YOLOv8InsightFaceOCRPython

Need this kind of work done?

See how engagements are structured, or get in touch directly.