▋AI-enabled data and software professional based in Jersey City, NJ. I build RAG and LLM workflows, analytics pipelines, predictive models, and backends in Python, SQL, and modern AI tooling—where data quality, retrieval, and production constraints meet.
Validation, ETL, retrieval, and guardrails—so models and dashboards stay grounded when stakes are high.
MS CS @ Montclair State (3.9); ex–Accenture & Infor backend work at real scale.
About
AI-enabled data and software professional with experience across generative AI workflows, data analytics, backend engineering, and applied machine learning. I've built and supported RAG-based systems, LLM-integrated decision-support workflows, semantic retrieval pipelines, predictive models, and analytics dashboards using Python, SQL, Power BI, PostgreSQL, and modern AI tooling.
Hands-on experience in data validation, ETL, reporting automation, API development, and production support, with a track record of improving data reliability, retrieval effectiveness, system performance, and operational decision-making. MS in Computer Science from Montclair State University (GPA 3.9); BTech in Electronics & Communication Engineering from Bharath Institute of Engineering and Technology, India.
Coursework & academic projects
Software Engineering, Algorithms & Data Structures, Operating Systems, Machine Learning, Computer Architecture. Projects include multi-agent AI systems (LLMs, RAG), an ML-based attendance system, and Android/mobile applications.
Tools I trust
Programming & scripting
- Python
- Java
- SQL
- JavaScript
Generative AI & LLMs
- RAG
- LLM integration
- Prompt engineering & templates
- Semantic retrieval & embeddings
- AI guardrails & response validation
- Structured JSON outputs
- LLM evaluation & feedback loops
- AI-assisted decision support
- Multi-agent systems (AutoGen)
Machine learning & NLP
- Logistic regression & classification
- Predictive modeling
- Feature engineering & threshold tuning
- Drift monitoring & PSI-based validation
- NLP
- Sentence Transformers & SBERT
- OCR (Tesseract)
- Rule-based systems
Data engineering & analytics
- ETL pipelines
- Data validation & quality checks
- Schema evolution & backfill recovery
- Data cleaning & transformation
- Query-based reporting & KPI tracking
- Ad hoc & trend analysis
- Reporting automation
Databases & storage
- PostgreSQL
- SQLite
- ChromaDB
Backend & software engineering
- RESTful APIs
- Spring Boot
- Microservices
- API contracts
- High-volume data processing
- Production support & incident resolution
- Performance optimization
BI & visualization
- Power BI
- Interactive dashboards
- Data visualization
- Performance & operational reporting
Cloud, DevOps & tools
- Jenkins
- Kibana
- Modal
- Gradio
- Git
- Pandas
- JUnit
- Mockito
Concepts & practices
- Data modeling
- Query optimization & indexing
- Schema validation
- Unit & integration testing
- CI/CD
- System reliability
- Stakeholder & cross-functional collaboration
Projects that ship
01
Pfizer Supply Chain Document Processing
RAG, OCR benchmarking, and LLM evaluation over pharmaceutical PDFs—externship project with a Gradio document chat UI.
LlamaIndex · FAISS · ChromaDB · PyMuPDF · Gradio · Extern
02
Multi-Agent Trading Platform with MCP
Four-agent trading simulation with AutoGen, market data, web search, SQLite portfolio state, and a Gradio monitoring dashboard.
Python · AutoGen · LLM integration · SQLite · Gradio
03
Autonomous Deal Detection System
Multi-agent pipeline over 100+ RSS feeds, ChromaDB retrieval at ~400K embeddings, and Modal deployment.
RAG · ChromaDB · SentenceTransformers · LLMs · Modal
Where I've built
Montclair State University
Montclair, NJ
Generative AI & Data Specialist
Aug 2024 – Present
- Owns ETL and schema evolution across Tutor.com, study hall, and recitation data—backfills, SLAs, and ~30% fewer cross-system inconsistencies on 10K+ weekly records for downstream LLM and analytics.
- Data validation, PSI-based drift monitoring, and a logistic regression early-warning model (ROC-AUC 0.63→0.81) tied into LLM-assisted student support.
- Shipped a RAG-based advisor workflow plus LLM evaluation, guardrails, and structured outputs for grounded academic interventions.
Data Analyst – Center for Academic Success and Tutoring
Feb 2024 – Aug 2024
- Consolidated and validated operational data across three teams (25K+ student records) for reporting and analysis.
- Built Power BI dashboards and recurring KPI reports for engagement, utilization, and team performance.
- Trend and gap analysis for staffing and scheduling; streamlined datasets to scale BI and later AI-ready analytics.
Accenture
Multi-region
Full Stack Engineer
Apr 2021 – Aug 2023
- Backend APIs and PostgreSQL for multi-region platforms (~5M users, 6+ countries); high-volume reporting with strong uptime.
- Query and index tuning (~70% faster execution; report load from minutes to under 10 seconds); SBERT semantic retrieval over 100K+ records wired into APIs.
- 200+ production incidents triaged with Kibana/PostgreSQL; translated 100+ requirements into schemas, validation, and API contracts.
Infor
Hyderabad, India
Junior Software Engineer
Oct 2020 – Mar 2021
- Spring Boot performance work: native indexed queries and pagination (~480ms→~310ms under load).
- OCR microservice (Python, Tesseract, logistic regression) with higher precision on labeled scans.
- ERP data integrity via JSON schema and constraints; JUnit/Mockito; Jenkins pipelines with fewer failed deploys.
Ask my AI Agent
Ask anything about my background, projects, or how I approach AI engineering problems. This agent is powered by the same tooling I use in production systems.
Examples
The agent is tuned to talk about Deepika's background, projects, and engineering decisions.
Let's build something real
Open to AI/ML, data engineering, and backend roles in the NYC metro area (Jersey City / remote-friendly).