Summary
AI Architect with 8+ years building and productionizing machine learning and Generative AI systems across healthcare, telecom, and finance. Career started February 2018 at Ericsson. Strong in LLM/RAG architecture, applied ML, and cross-functional delivery. M.Tech (Data Science) candidate at IIT Hyderabad.
Experience
AI Researcher — LightSpun · Healthcare · Jan 2024 – May 2025
Built healthcare Generative AI and agentic workflows with RAG/LLM grounding, safety constraints, and production handoff in regulated settings. Applied RAG and LLM patterns to clinical and operations knowledge use cases with evaluation and guardrails.
Senior Data Scientist — IRIS Software · Finance · Jan 2024 – Present
Building end-to-end Credit Risk solution for Bank of Montreal using DL, GenAI over AWS integrating GCP models. Designed robust MLOps platform over AWS to operate thousands of models while supporting 300+ data scientists.
Lead Data Science — Bharti Airtel · Network · May 2022 – Jan 2024
Network analytics and ML leadership at telco scale — planning and performance decision systems.
Engineer, Machine Learning — L&T Technology Services · Feb 2021 – May 2022
ML engineering, applied modeling, Intel frameworks, production-minded pipelines.
Data Scientist — Ericsson · Feb 2018 – Jan 2021
Career start. Time series (ARIMA), classical ML, and applied data science for telecom domains.
Education
M.Tech, Data Science — Indian Institute of Technology Hyderabad · June 2024 – Present · 7.5/10
B.Tech, ECE — JIS College of Engineering · Aug 2013 – July 2017 · 8.5/10 · Top 1% · Best undergraduate researcher
Certifications & programs
- Generative AI with Large Language Models (Sept 2023)
- Master's Program: AI Engineering — Simplilearn (2019–2020)
- AWS Machine Learning Nanodegree — Udacity (Oct 2021)
- Internet of Things XEE100-017 — Stanford University (Jan 2021)
Skills
GenAI / Agents: LLM system design, RAG / GraphRAG, multi-agent orchestration, MCP, A2A, Google ADK, LangGraph, LangChain, Vertex AI Agent Engine, OpenAI Agents / Responses patterns, Anthropic tool & computer-use patterns, agent memory (session / episodic), browser / computer-use agents, Agent Skills (SKILL.md), structured generation, prompt & context engineering, evals (LangSmith / Braintrust / DeepEval, online + offline in CI), tool-layer security, durable / long-running agents, semantic & prompt caching, guardrails, LLMOps, Gemini/GPT/LLaMA/BERT, PEFT / fine-tuning, document AI / multimodal
ML / Data: PyTorch, TensorFlow/Keras, Hugging Face, scikit-learn, deep learning, computer vision, NLP, time series, self-supervised / anomaly detection, Pandas/NumPy/Dask/PySpark
Languages / APIs: Python, TypeScript, SQL, Bash, FastAPI, GraphQL, Next.js/React
Cloud / Platform: GCP Vertex AI, Cloud Run, BigQuery, Firestore, GCS, Pub/Sub, Cloud Tasks, AWS SageMaker, Docker, Kubernetes, Cloud Build, Pulumi, PostgreSQL/Cloud SQL, Redis, MLflow, Looker/Tableau
Architecture practice: reference architecture, cost/latency/quality trade-offs, human-in-the-loop, data contracts, PHI/PII & security, tool auth / least-privilege / agent identity, prompt-injection defense at tools, multi-tenant RBAC, durable execution (workflows / queues / Temporal patterns), OpenTelemetry, production runbooks