Software Engineer (AI/ML) • LLM Systems • Reliability • RAG
I build AI systems that actually work in the real world ⚙️ (not just on benchmarks)
- 🧠 Building a Confidence-Guided LLM Routing System (Thesis)
- ⚙️ Learning System Design, APIs, FastAPI, Docker
- 📊 Exploring LLM Evaluation, Calibration & Reliability
- 💻 Practicing DSA + Software Engineering fundamentals
-
🔹 LLM Routing System
→ Dynamic routing between small & large models using uncertainty -
🔹 Text-to-SQL (RAG System)
→ Natural language → SQL using FAISS + FastAPI -
🔹 LLM Calibration & Reliability Analysis
→ Fixing confidence–accuracy mismatch (ECE, temp scaling) -
🔹 Adversarial LLM Evaluation
→ Testing robustness & persona drift in language models -
🔹 Financial Risk Prediction System
→ ML pipeline with ROC-AUC up to 0.97
Languages: Python
ML/DL: PyTorch · Transformers · scikit-learn
LLM Systems: RAG · Prompt Engineering · LoRA
Backend: FastAPI · REST APIs
Infra: FAISS · SQLite · (learning Docker 🐳)
Data: Pandas · NumPy
- LLM Systems & Multi-model Architectures
- Model Reliability & Failure Analysis
- Cost vs Performance Trade-offs in AI
- Building real-world ML systems end-to-end
- 🏎️ Formula 1 enthusiast (race weekends > everything)
- ⚽ Football buff
- 🎮 Competitive & story-driven games
- 🎧 Music to reset the brain
💼 LinkedIn
📧 [email protected]
Precision matters — whether in model calibration or a last-lap overtake.