About Me
AI Engineer with hands-on expertise in Generative AI, LLMs, and scalable ML systems, delivering production-ready solutions across real-world use cases. Built and deployed RAG pipelines, NL-to-SQL systems, and AI-driven automation workflows with measurable impact on efficiency and performance. Strong in Python, FastAPI, LangChain, and vector databases, with deep understanding of NLP, data pipelines, and model fine-tuning. Focused on building robust, high-performance AI systems that bridge research and real-world deployment.
Skills & Technologies
Work Experience
AI/ML Engineer
Delivered 4+ production-grade AI systems serving 200+ users, improving efficiency and reducing manual effort across enterprise workflows. Designed and deployed Agentic AI workflows using LangChain and LangGraph, including a Natural Language-to-SQL system enabling non-technical users to query databases with 90%+ accuracy. Built and optimized RAG-based GenAI systems for enterprise knowledge retrieval across unstructured data (PDFs, CSVs), improving retrieval relevance by 60% and reducing hallucinations. Developed scalable AI services using FastAPI, supporting high-concurrency workloads and seamless integration with enterprise applications. Led end-to-end development of LLM solutions, including data preprocessing, embedding pipelines, LoRA fine-tuning (DeepSeek 1.5B), evaluation, and deployment. Automated enterprise workflows such as SOP and protocol generation, improving operational efficiency by 40%. Built and maintained data ingestion and transformation pipelines to ensure high-quality data flow for AI/ML systems. Implemented logging, monitoring, and evaluation frameworks to ensure performance, reliability, and stability of production AI systems.
Projects
Drone Security Agent (Agentic AI + Robotics Integration)
Overview: Developed an AI-powered drone control system using natural language commands, integrating LLM-based agents with real-world robotics systems for intelligent and safe drone operations. Key Contributions: ● Built an Agentic AI system using LangChain to translate natural language into actionable drone commands ● Integrated LLM reasoning with system-level controls for autonomous decision-making ● Implemented safety checks to validate commands before execution Impact: ● Demonstrates real-world AI + robotics integration ● Showcases ability to move beyond software into physical system control ● Relevant for automation, defense, and disaster response use cases Tech Stack: LangChain, Python, ROS concepts, LLMs
Product Intelligence System (GAN + LLM Fine-Tuning)
Overview: Built an advanced AI system combining generative models and LLMs to extract product insights and enhance decision-making. Key Contributions: ● Applied LLM fine-tuning concepts for domain-specific intelligence ● Integrated generative models (GAN-based concepts) with LLM pipelines ● Designed pipelines for extracting structured insights from unstructured product data Impact: ● Demonstrates multi-model AI system design ● Highlights ability to work with advanced architectures beyond basic ML ● Relevant for analytics, recommendation systems, and business intelligence Tech Stack: Python, LLMs, GAN concepts, NLP
Email Classification System (ML + NLP)
Overview: Developed a machine learning-based system to automatically classify emails into categories using NLP techniques. Key Contributions: ● Performed text preprocessing and feature extraction ● Built classification models for multi-class categorization ● Evaluated performance using standard ML metrics Impact: ● Automates email triaging and workflow management ● Reduces manual effort in communication-heavy systems ● Strong example of applied NLP in business workflows Tech Stack: Python, Scikit-learn, NLP
News Summarization with Text-to-Speech (GenAI + UX)
Overview: Built a system that summarizes news articles using LLMs and converts them into audio using text-to-speech. Key Contributions: ● Implemented LLM-based summarization pipeline ● Integrated text-to-speech (TTS) for audio output ● Designed end-to-end pipeline from ingestion → summarization → delivery Impact: ● Enhances content accessibility and consumption ● Useful for news aggregation, accessibility tools, and productivity apps ● Demonstrates multi-modal AI capabilities (text + audio) Tech Stack: Python, LLM APIs, TTS systems
Resume Categorizer (NLP Classification)
Overview: Built an NLP-based system to classify resumes into job categories. Key Contributions: ● Implemented ML models (KNN, OneVsRest) ● Designed preprocessing pipelines for text normalization ● Achieved high classification accuracy across multiple categories Impact: ● Automates HR screening workflows ● Improves efficiency in recruitment pipelines ● Demonstrates practical use of ML in business operations Tech Stack: Python, Scikit-learn, NLP
MNC Promotions Data Analysis (Data Analytics)
Overview: Performed exploratory data analysis to identify key factors influencing employee promotions. Key Contributions: ● Conducted EDA and bivariate analysis ● Identified trends in performance, recruitment channels, and growth patterns ● Generated actionable insights from structured datasets Impact: ● Supports data-driven HR decision-making ● Demonstrates ability to extract insights from business data ● Shows foundation in analytics and statistical reasoning Tech Stack: Python, Pandas, NumPy, Data Visualization