Portfolio © 2026 — AI / ML Engineer
Mohammed Rabbani.
Final-year B.Tech AIML student crafting intelligent systems — from adaptive RAG pipelines to real-time computer vision. I care about work that ships, not notebooks that sit.
About
I'm a final-year B.Tech student in Artificial Intelligence & Machine Learning at Biluru Gurubasava Mahaswamiji Institute of Technology — passionate about building AI systems that actually ship, not just notebooks.
My work spans retrieval-augmented generation, computer vision with explainability, swarm optimization research and multi-agent systems. I care about honest evaluation, resource-efficient engineering — most of my projects run on free-tier constraints — and deploying models people can try in the browser.
Selected work
(06)- 01 Adaptive-RAG Query-aware routing between sparse, dense & multi-hop retrieval — 65% faster than fixed-hybrid on ArguAna, built entirely on a $0 stack. ↗
- 02 Plant Disease Detector 38 plant diseases from leaf images at 99.1% validation accuracy — ResNet18 transfer learning with Grad-CAM explainability. ↗
- 03 SleepGuard Browser-based drowsiness detector — FaceMesh eye tracking, multi-channel alarm, zero backend. Works over any app via Picture-in-Picture. ↗
- 04 EcosystemBridge Apple-style continuity between Windows and Android — your PC and phone working in tandem. ↗
- 05 UAV Swarm Interception Hybrid GWO–LSTM framework — predicting evasive target trajectories, then optimizing swarm interception paths in a closed loop. ↗
- 06 ShadowIntern Multi-agent concept that studies a company's engineering culture, then generates mock tasks, code reviews & interview simulations. ↗
Capabilities
- Languages Python · JavaScript · SQL
- ML / Deep learning PyTorch · TensorFlow / Keras · scikit-learn · Transformers · FAISS · BM25
- Data Pandas · NumPy · SciPy
- Tools & deployment Streamlit · Hugging Face · Git · MediaPipe · Web APIs
Education
B.Tech — Artificial Intelligence & Machine Learning
Biluru Gurubasava Mahaswamiji Institute of Technology
Contact
Let's build something*
Open to internships, research collaborations and interesting AI/ML problems — my inbox is always open.