Mohammed Syed

About

Backend/ML engineer focused on clinical AI transparency who built ShifaMind, an explainability-enforced clinical ML system using UMLS/MIMIC-IV/PubMed data with RAG, GraphSAGE, and cross-attention. Demonstrated strong production engineering via FastAPI API design and safe migrations (feature flags/shadow inference), plus HIPAA-aligned auth/RLS patterns; also delivered a real-time comet detection system reaching 97.7% accuracy.

Experience

Education

Key Strengths

• Designed explainability-first clinical ML architecture (not post-hoc) using concept grounding and alignment
• Improved model performance vs baseline (79.89 Macro F1, +11.1%; 91.55 AUROC) while maintaining explainability
• Adapted TCAV for concept–diagnosis relationships; achieved 0.904 average concept predictability score and concept completeness 1.0
• Built scalable ML-backed APIs with FastAPI using schema validation (Pydantic), versioned endpoints, and isolated inference orchestration
• Scaled compute-heavy inference via Docker + RunPod with caching and batching
• Led safe backend refactor/migration using feature flags, shadow inference, staged rollout, and rollback-safe immutable artifacts
• Implemented production auth and access controls with Firebase Auth + database-enforced row-level policies for HIPAA-aligned systems
• Identified and fixed overlooked edge cases in imaging pipeline, improving real-time detection accuracy to 97.7%