Data & AI Systems Professional
Designing scalable analytics platforms and human-centered AI solutions for business impact.
Education
University of Maryland
Master of Science, Information Systems
PES University
Bachelor of Technology, Electronics and Communication
Skills Snapshot
Python, SQL, R, Polars, Spark, Snowflake, HiveQL, Tableau, Power BI, Excel
LLM Evaluation, Prompt Engineering, RAG, Neo4j, N8N, HuggingFace Transformers, LoRA Fine-tuning (Gemma), Streamlit
AWS, GCP, Azure, Docker, HDFS, Git, CI/CD, CRON
Microsoft Project, Feasibility Analysis, Functional Modeling, Process Documentation
Experience
Research Assistant
R.H. Smith School of Business · University of Maryland
- Optimized Snowflake SQL workflows and engineered Polars batch pipelines, reducing end-to-end processing for 100M+ job market records from 2 hours to 15 minutes.
- Curated and transformed domain-specific datasets for LLM fine-tuning, improving downstream model performance and context relevance.
- Designed and implemented LLM evaluation pipelines benchmarking GPT-4, Claude, Llama-3, and Qwen across 10K+ job postings; insights informed academic research and CNN/WSJ-featured reports.
- Built ETL systems integrating five disparate data sources to support the AI Jobs Dashboard for labor market research.
System Engineer
Oracle Cerner · Bengaluru
- Resolved critical Linux application and Sybase database issues for hospital systems, reducing incident tickets from 20 to 5 per week while maintaining 24/7 availability.
- Collaborated with client IT teams to troubleshoot server incidents, achieving a 75% improvement in mean resolution time.
SAP Consultant
Accenture · Bengaluru
- Analyzed and resolved 500+ SAP issues, maintaining 95% SLA compliance for supply chain operations.
- Created Standard Operating Procedures during a ServiceNow migration, reducing process violations by 80%.
Projects
Graph Search MCP
AI-ready knowledge graph platform
- Built an AI-ready knowledge graph platform by integrating Neo4j with LLMs via the Model Context Protocol (MCP); containerized with Docker for consistency and portability.
- Fine-tuned Gemma 3n using LoRA optimization on relationship data, improving relation extraction accuracy.
- Developed hybrid graph + LLM retrieval systems that combined symbolic traversal with neural reasoning, achieving a 73% lift over vector-only baselines.
Asha IT Recruitment Platform
AI-driven hiring workflows for small businesses
- Architected an AI recruitment platform using AWS Amplify, Lambda, and DynamoDB, integrating Firebase Authentication for secure user management.
- Implemented agentic AI pipelines that matched candidates to roles with LLM-based skill extraction, resume–JD reasoning, and multi-agent ranking.
- Optimized Lambda-based inference latency by 42% through asynchronous invocation patterns and caching layers.
Dynamic Surge Pricing
Predictive parking demand with PySpark
- Built a PySpark regression model on Seattle parking data to predict availability and apply surge pricing rules for 2023 live streams.
- Containerized the pipeline with Docker, deployed on Vertex AI, and automated ingestion via Pub/Sub with results in BigQuery.
- Visualized surge pricing outputs on Streamlit and monitored reliability through Cloud Logging.
Leadership & Recognition
Smith Master's IS Association · VP Finance
Jan 2025 – Present · Developed a mentorship program connecting 100+ students and alumni; featured for AI-driven engagement.
University ArticleAI in Business Case Competition Winner
Nov 2024 · Designed an AI integration strategy enabling small businesses to automate operations and customer insights.