MS Computer Science @ UW-Madison · Ex-SDE @ Accenture · AI/RAG Engineer · Published Researcher
Google Scholar •
LinkedIn •
Email
I'm an MS CS student at UW-Madison (GPA: 4.0) and former Advanced App Engineering Analyst at Accenture, where I built production-grade RAG pipelines and AI systems for financial data at scale. My work sits at the intersection of AI agents, distributed systems, and MLOps.
- 🎓 MS Computer Science @ University of Wisconsin-Madison (2025–2027)
- 💼 Ex-Accenture: built RAG pipelines cutting pipeline runtime from 5 days → 24 hours (80% improvement)
- 🔭 Currently working on distributed systems, foundation models, and real-time data pipelines
- 🌱 Exploring LLM tooling, Kafka streaming, and full-stack AI apps
- 📄 Published researcher — 3 papers, 8+ citations
- 🌍 Exchange student — University of Tours, France (2021–2022)
| Project | Description | Stack |
|---|---|---|
| ⚡ transaction-feed | Real-time Kafka-style transaction monitoring via WebSockets, 200+ concurrent users, sub-500ms latency | React · TypeScript · Node.js · WebSocket |
| 💰 finance-dashboard | Full-stack expense tracker with Recharts visualizations and sub-200ms response times | React · TypeScript · Flask · PostgreSQL |
| 🌤️ weather-streaming | Distributed Kafka pipeline — 10 msg/sec across 4 partitions, Protobuf serialization, exactly-once processing | Python · Kafka · gRPC · Docker |
8 projects covering the full modern data engineering stack:
| Project | Topic | Key Technologies |
|---|---|---|
| P1 — Dockerized Git Analyzer | Dockerized tool to clone, diff, and LLM-summarize Git branches | Docker · Shell · LLM |
| P2 — gRPC Multi-Container App | Fault-tolerant multi-container gRPC + HTTP service with caching | gRPC · Docker · Protobuf · Flask |
| P3 — Concurrent Word Count & Benchmarking | Multi-threaded word counter with format benchmarking | Python · Parquet · Apache Arrow · Concurrency |
| P4 — SQL & HDFS Pipeline | Fault-tolerant SQL → HDFS data pipeline via gRPC and WebHDFS | HDFS · SQL · gRPC · PyArrow · Docker |
| P5 — Spark & Hive Analysis | Spark + Hive analysis of competitive programming dataset | Apache Spark · Hive · PySpark · Docker |
| P6 — Cassandra Weather Pipeline | gRPC weather data ingestion into Cassandra, consistency tradeoffs | Cassandra · gRPC · Protobuf · Docker |
| P7 — Kafka Weather Streaming | Kafka producer/consumer pipeline for weather station data | Kafka · Python · Streaming · Docker |
| P8 — GCP BigQuery Analysis | GCS + Dataform + BigQuery pipeline for Wisconsin schools dataset | GCP · BigQuery · Dataform · SQL |
Advanced App Engineering Analyst @ Accenture · Sep 2023 – Aug 2025
- Built distributed RAG ingestion framework (FastAPI, OpenAI, Milvus, PostgreSQL) — reduced runtime 5 days → 24 hours (80%) with 95% query accuracy
- Led Azure DevOps CI/CD orchestration saving 100+ analyst hours/month
Data Engineering Intern @ National Institute of Ocean Technology · Jul 2022 – Jun 2023
- MPC algorithm for OTEC in MATLAB — reduced steady-state error by 25%, improved transient response by 18%
- LSTM model for disturbance rejection improving system stability by 30%
SDE Intern @ Carjoz · Sep 2021 – Jan 2022
- Android app supporting 200+ daily records, Firebase optimization for 50+ active users
Languages
AI / ML
Data & Distributed Systems
DevOps & Cloud
Web
| Degree | Institution | GPA | Year |
|---|---|---|---|
| MS Computer Science | University of Wisconsin-Madison | 4.0 | 2025–2027 |
| BE Electronics & Instrumentation | MIT Campus, Anna University | 3.87 | 2019–2023 |
| Exchange Program | University of Tours, France | 3.92 | 2021–2022 |
| Paper | Venue | Year | Citations |
|---|---|---|---|
| Decentralized Control Strategies with Predictive Disturbance Rejection for OC-OTEC Plant | Journal of Water Process Engineering | 2024 | 2 |
| Multiple Hybrid AI Model-based Optimal Charging Profile of Li-ion Battery for SOH Enhancement | IEEE | 2023 | 4 |
| Data-driven Model Development and Validation for Laboratory Scale OC-OTEC Plant | IEEE | 2022 | 2 |
