Experience

Machine Learning & AI

Creating intelligent systems that learn and adapt

I have experience building LLM-powered applications, ML pipelines, and AI systems. From RAG architectures to production ML deployment, I love working on intelligent solutions.

Work Experience

Summer Engineer Analyst

Goldman SachsDallas, Texas, USAAug 2025

Built and deployed TalentFlowAI, a multi-step agentic LLM pipeline using LangChain and tool calling to automate candidate profiling
Implemented embedding-based semantic retrieval and ranking pipelines to automate applicant prioritization and recruiter screening workflows
Designed decision orchestration logic to generate screening summaries and recommend next-stage actions using confidence signals
Developed recruiter-facing web interfaces in React and TypeScript to surface AI outputs with traceability and human-in-the-loop controls
Optimized Spring Boot services and MongoDB pipelines with asynchronous processing, improving inference and retrieval latency by 5%
Built reusable, WCAG-compliant UI components aligned with GS design system to standardize AI-powered recruiting workflows

Product Manager, Research Systems

Saint-Gobain ResearchTamil Nadu, IndiaJun 2024

Built a data analytics platform using regression model workflows to model experimental material performance across 15 research programs
Designed standardized data schemas and metadata pipelines to enable large-scale comparison of 30+ material scenarios
Developed Node.js services and Python pipelines for automated feature extraction, validation, and data quality monitoring
Integrated research datasets with Salesforce and HubSpot APIs to enable ML-driven technical reporting and customer insights

Associate Product Engineer

TeleAppsTamil Nadu, IndiaMar 2023

Built RESTful ML analytics services using TensorFlow and Flask to power customer engagement and campaign performance platforms
Deployed containerized ML microservices for real-time inference and internal reporting, implementing model versioning, lazy loading, and request-level caching, reducing cold-start latency and inference load time by ~20%
Created end-to-end feature engineering pipelines using pandas and time-series transformations including lag features, rolling windows, seasonality indicators, and trend decomposition, improving model stability by ~15%

Technology Analyst

Bahwan CyberTekChennai, Tamil Nadu, IndiaFeb 2022

Supported enterprise data by extracting, validating, and reconciling large datasets across multiple systems to improve client reporting accuracy
Built Python based data analysis scripts and dashboards to evaluate user behavior and market trends for consulting and delivery teams
Assisted in configuring project workflows using Adobe Workfront, supporting task tracking, documentation, and cross team coordination across analytics projects

Technical Skills

Large Language Models (GPT-based, OpenAI API, HuggingFace)Prompt EngineeringTool and Function CallingRetrieval-Augmented Generation (RAG)Embedding ModelsSemantic RetrievalAutonomous ReasoningDecision OrchestrationVerification LoopsHuman-in-the-Loop SystemsSupervised and Unsupervised LearningEnsemble MethodsXGBoostLightGBMRandom Forestscikit-learnTensorFlowPyTorchFeature EngineeringCross-ValidationHyperparameter OptimizationAnomaly DetectionModel Explainability (SHAP)Distributed Data PipelinesETL and Feature StoresBatch and Streaming ProcessingRESTful and Async APIs (FastAPI, Flask, Spring Boot)Vector Databases (FAISS)MongoDBPostgreSQLRedisCaching and Query OptimizationWorkflow OrchestrationContainerized ML Services (Docker)Model Versioning and Experiment Tracking (MLflow)CI/CD PipelinesScalable Inference ArchitecturesAPI GatewaysMonitoring and LoggingDrift DetectionPerformance ProfilingCloud Platforms (AWS, GCP)

Related Projects

View All

ML/AI

TaskPilot AI

Autonomous Workflow Execution System

An AI workflow engine using OpenAI APIs, FAISS vector search, and FastAPI to interpret and execute 1,000+ multi-step task scenarios

Mar 2025

ML/AI

HireSense AI

Resume & Interview Intelligence Platform

An AI-powered recruiting intelligence platform using HuggingFace Transformers, OpenAI embeddings, Pinecone, and RAG pipelines

May 2024

ML/AI

RiskGuard ML

Transaction Fraud Detection System

Supervised and semi-supervised fraud detection models using XGBoost, Random Forest, and PyTorch on 5M+ transactional records

Feb 2024