AI Engineer with 1+ years in Generative AI & Data Science
AI is analyzing your overall score…
Identifying your key strengths…
Evaluating your skill match against the job requirements…
Assessing your cultural and operational fit
Highly motivated and results-driven AI Product Developer with 1.0 years of experience in building and deploying AI-driven pipelines for product search, structured extraction, and fraud detection. Proficient in machine learning, deep learning, generative AI, and full-stack development, with a strong background in data analytics and scalable systems. Eager to apply analytical and technical skills to develop innovative AI solutions.
Vellore Institute of Technology, Chennai
M.Tech (Integrated) · CSE with Business Analytics
August 1, 2022 – June 30, 2027
Alfaleus Tech
Software Development Intern
June 1, 2025 – July 1, 2025
India
BAE.AI
AI Product Developer
May 1, 2025 – Present
Singapore
CODIIS
Data & ML Intern
June 1, 2024 – August 1, 2024
India
Agentic AI Meeting & Workflow Assistant
September 1, 2024 – Present
• Designed and built a full-stack meeting intelligence system with an 8-endpoint RESTful FastAPI backend, handling transcript ingestion, LLM extraction, PostgreSQL persistence, and Markdown export • Engineered a 2-pass agentic LLM pipeline using Gemini Flash that extracts structured JSON (summaries, decisions, action items, assignees) from raw transcripts — with a deterministic fallback validation layer to prevent hallucinated assignees from entering the workflow • Implemented PostgreSQL full-text search across meeting titles, summaries, and raw transcripts using SQLAlchemy, with schema versioning via Alembic migrations • Built persisted, context-aware chat per meeting, maintaining conversation history in the database to enable multi-turn Q&A grounded in meeting content • Reduced manual post-meeting overhead by automating follow-up email drafting, action item extraction, and assignee resolution – all via a single /api/meetings/process call
AI Powered Resume Matching & Job Aggregator
January 1, 2024 – May 1, 2024
• Built a two-stage retrieval and LLM re-ranking pipeline, surfacing Top-5 internship matches with structured scoring • Engineered automated job-ingestion pipelines with fallback extraction, deduplication, and Firestore-backed storage • Developed resume parsing and semantic matching workflows through a Streamlit interface • Automated scheduled data-refresh pipelines using GitHub Actions for continuously updated opportunity retrieval
Kaarigar Konnect – AI Powered Full-Stack Marketplace
January 1, 2024 – April 1, 2024
• Built an AI-powered e-commerce platform with onboarding, product listings and order management on Firebase • Architected a multimodal Voice-to-Listing pipeline converting voice input into structured product metadata • Implemented generative AI for image generation and embedding-based semantic retrieval over Firestore • Built a multilingual inference pipeline with translation caching and zero-reload UI updates • Integrated Gemini text-to-speech with custom audio encoding for browser playback
Context-Aware Document Q&A Chatbot
November 1, 2023 – February 1, 2024
• Built a zero-backend RAG pipeline entirely in the browser: client-side PDF parsing via pdf.js web workers → chunking (300-word segments, 50-word sliding overlap) → TF-IDF vectorization → cosine similarity retrieval – eliminating server infrastructure and API data exposure • Implemented an in-browser retrieval engine that builds a global vocabulary across all uploaded documents and selects the top-5 most relevant chunks per query, injecting them into a strict Gemini 1.5 Pro prompt for grounded, citation-backed answers • Designed a structured JSON response schema for the LLM output that includes the answer, confidence score, and primary source citation – surfaced in the UI with color-coded confidence indicators and collapsible source cards • Engineered out-of-scope detection at the prompt level, preventing hallucinations by instructing the model to explicitly flag queries not answerable from the uploaded documents
Trackflow - Full-Stack CRM & Analytics Dashboard
September 1, 2023 – November 1, 2023
• Built Supabase Realtime subscriptions for sub-second UI sync across 6-stage lead and 4-stage fulfillment pipelines • Reduced manual reporting by centralizing 3+ KPIs into a Chart.js dashboard with dynamic time filtering • Eliminated order communication lag via automated emails on fulfillment state changes using Supabase Edge Functions • Built PostgreSQL CTEs and window functions for funnel drop-off and SLA breach analysis, replacing ad-hoc reporting
AML Network Detection – Quantum Graph Optimization
June 1, 2023 – August 1, 2023
• Formulated AML detection over 1.32M transactions as a QUBO using D-Wave Systems Ocean SDK • Designed a constrained fraud-optimization objective, achieving ROC-AUC 0.9764 and F1 0.7302 without labels • Recovered 89 suspicious ego subgraphs from a transaction graph, identifying 15.1 fraudulent edges per alert on average
Real-Time Fraud Detection – Graph & Streaming Analytics
April 1, 2023 – June 1, 2023
• Simulated a high-throughput Kafka producer at 1,000+ transactions/sec with configurable partition scaling • Achieved sub-100ms end-to-end latency from event ingestion to fraud flag and reduced false positive rate by 23% over baseline • Used Neo4j Cypher for sub-10ms multi-hop queries across 3 node types, enabling real-time transaction decisions • Eliminated batch retraining via River’s online learning, updating models per event with zero downtime
Glaucoma Early Detection – Hybrid Quantum-Classical ML Pipeline
February 1, 2023 – April 1, 2023
• Built a gene expression analysis pipeline with DEG selection, FDR correction, and stratified cross-validation • Implemented leakage-free within-fold feature selection for robust high-dimensional biomedical classification • Designed a hybrid variational quantum classifier in PennyLane using PCA-reduced inputs and 4-qubit circuits • Outperformed classical baselines in F1-score (0.833 vs 0.800), improving minority-class glaucoma detection
BNPL Consumer Behaviour Study
December 1, 2022 – February 1, 2023
• Designed and administered a primary survey (n=60) across Gen Z consumers to study BNPL adoption behaviour, operationalizing 8 constructs including impulsive buying, gratification, spending control, and peer influence • Applied Technology Acceptance Model (TAM) and Uses & Gratifications Theory (UGT) as theoretical frameworks; performed Cronbach's alpha reliability testing across all constructs to validate internal consistency of survey scales • Analyzed relationships between psychological, technological, and social factors driving BNPL adoption producing actionable insights relevant to fintech product design and financial behavior modelling • Conducted structured statistical analysis on consumer spending patterns with implications for credit risk, financial inclusion, and digital payment ecosystem design
Financial Market Analytics Platform
October 1, 2022 – December 1, 2022
• Built a realtime stock data streaming pipeline using Apache Kafka with a producer-consumer architecture, ingesting live price data at configurable intervals • Applied time series decomposition (trend, seasonal, residual) using statsmodels and Linear Regression to predict stock price highs and lows • Developed hybrid LSTM + Weighted Moving Average model augmented with TextBlob news sentiment analysis on RELIANCE.NS financial data • Conducted deviation analysis comparing model predictions against actual closing prices and proposed improvements using FinBERT and ARIMA ensemble methods
Cisco Networking Academy – Introduction to Packet Tracer (Networking & IoT Simulation)
Cisco Networking Academy
June 1, 2026 – Present
Cultural Fit Analysis
The candidate's diverse range of personal projects, spanning various domains like finance, healthcare, e-commerce, and general AI applications, demonstrates a strong curiosity and willingness to explore different challenges. This breadth of interest, combined with experience in both academic (CODIIS intern) and startup (BAE.AI, Alfaleus Tech) environments, suggests adaptability and a proactive learning mindset, which are positive indicators for cultural fit in a dynamic AI engineering team.
Soft Skills & Operational Fit
The candidate's project descriptions indicate strong problem-solving skills, evidenced by engineering solutions for hallucination prevention, cost optimization, and real-time performance. Their ability to design and implement complex systems (e.g., agentic LLM pipelines, real-time fraud detection) suggests a proactive and detail-oriented approach. The documentation and presentation experience at Alfaleus Tech also points to good communication and collaboration potential.