Building production ML, full-stack GenAI, and agentic AI systems end-to-end
American Airlines · Amazon · MS in Data Science, Indiana University

AI/ML engineer with 6+ years shipping production ML, full-stack GenAI, and agentic AI systems and AI/ML platforms end-to-end at American Airlines and Amazon — from ambiguous business problems through data, models, agents, applications, evaluation, and production operations.

• 🛫 Built AI/ML solutions with a $100M/yr revenue impact across 100+ pricing routes at American Airlines
• 🤖 Built a RAG multi-agent assistant serving 3M+ requests/day, delivering $6M+/yr in cloud savings at 99.9% serving SLOs
• 📉 Cut hallucination 14% → 4%, p95 latency 2.1s → 780ms, inference cost $6.80 → $2.10 / 1K queries
• 🎓 MS in Data Science, Indiana University Bloomington · BTech, NIT Nagpur

Agentic AI / GenAI

ML / DL Frameworks

Serving · GPU · MLOps

Cloud · Data · Infra

Apps · APIs · Languages

Developer, AI/ML — Pricing · American Airlines · 09/2024 – Present Own AI/ML solutions, agents, platform services, and production operations behind Pricing & Revenue Management across 100+ routes ($100M/yr optimization value). Agentic pricing analysis with governed MCP tools, tiered autonomy, and HITL approvals; production evaluation/governance platform for agents and models.

Software Engineer — GenAI, Agents & ML/AI Platforms · Amazon · 08/2022 – 09/2024 Architected and scaled a conversational RAG assistant at 3M+ req/day on a $50M-revenue surface. In-house ReAct agent runtime, multi-agent orchestration, LoRA/SFT/DPO fine-tuning, two-tower/DLRM ranking (CTR +9%, conversion +6%).

Machine Learning Engineer · Flux Auto · 01/2018 – 12/2020 Perception & classification pipelines with repeatable MLOps workflows, slice-based evaluation, automated data-quality checks, and retraining via hard-example mining.

📫 Reach me: [email protected] · LinkedIn · Dallas–Fort Worth, TX