Professional Summary
Results-driven AI/ML Engineer with around 3 years of experience designing, developing, and deploying enterprise-grade Generative AI,
Agentic AI, and Machine Learning systems for production environments. Strong expertise in Python, FastAPI, LangChain, LangGraph,
CrewAI, LLMs, and Retrieval-Augmented Generation (RAG) for scalable backend systems. Hands-on experience across the full ML
lifecycle — data preprocessing, feature engineering, model training and evaluation, and production deployment — for classification,
regression, and computer vision tasks using Scikit-learn, PyTorch, and TensorFlow. Experienced in building multi-agent AI systems,
AI workflow automation, voice-enabled AI assistants, intelligent document processing, and enterprise chatbot solutions, as well as
fine-tuning open-source LLMs using LoRA and QLoRA and optimizing inference performance through prompt engineering and caching
techniques. Skilled in integrating vector databases (Pinecone, Qdrant, FAISS, Milvus, ChromaDB) for semantic search and enterprise
knowledge retrieval, with deployment experience across Docker, AWS, and CI/CD pipelines.
Technical Skills
| Programming Languages | Python, JavaScript (Basic), Bash, HTML |
| Generative AI | LangChain, LangGraph, CrewAI, OpenAI SDK, GPT-4, Claude, Gemini, LLaMA, BERT, T5, Prompt Engineering |
| Agentic AI | Multi-Agent Systems, MCP (Model Context Protocol), Tool Calling, Workflow Orchestration, Memory Management, Human-in-the-Loop |
| RAG & Vector Databases | Pinecone, Qdrant, FAISS, Milvus, ChromaDB, Neo4j |
| Backend Development | FastAPI, Flask, REST APIs, Microservices, WebSockets |
| Voice AI | Twilio Voice, Deepgram, ElevenLabs, OpenAI STT, Edge TTS, AssemblyAI |
| Machine Learning | Feature Engineering, Model Development, Model Deployment, MLOps, Scikit-learn, PyTorch, TensorFlow, Hugging Face Transformers, CNN, RNN, LSTM |
| Computer Vision | OpenCV, ResNet, Image Classification, Object Detection |
| Concurrency & Performance | Asyncio, Async/Await, Multithreading, Multiprocessing, ThreadPoolExecutor, ProcessPoolExecutor, Celery |
| Automation & Scraping | n8n, Zapier, Zoho, Selenium, Playwright, BeautifulSoup, OCR |
| Databases | PostgreSQL, Redis |
| Cloud & DevOps | Docker, AWS (EC2, S3), Git, GitHub, CI/CD, Linux, DVC, MLOps |
Professional Experience
AI/ML Engineer— Suffescom Solutions Pvt. Ltd.
Mar 2026 – Jun 2026
Python, FastAPI, LangChain, LangGraph, CrewAI, MCP, Pinecone, ChromaDB, Docker, AWS, LoRA, QLoRA
- Designed and developed enterprise-grade Agentic AI applications using LangGraph and CrewAI to automate complex business workflows.
- Implemented MCP (Model Context Protocol) for standardized tool calling, enabling agents to securely discover and invoke external tools and enterprise APIs; deployed to production using Docker and AWS.
- Built end-to-end RAG pipelines and fine-tuned open-source LLMs using LoRA/QLoRA; developed scalable backend services with FastAPI and Docker deployment.
- Developed AI-powered web automation, intelligent document processing, and data extraction solutions using Python and enterprise automation tools.
- Owned the end-to-end ML lifecycle — feature engineering, model development, and production deployment — following MLOps best practices for reproducibility and monitoring.
- Optimized AI workflows through prompt engineering, vector search, and caching strategies, improving response accuracy and system performance.
- Collaborated with cross-functional teams to deliver production-ready AI solutions following Agile development practices.
Associate AI/ML Engineer— Eminence Internet Technology Pvt. Ltd.
Apr 2024 – Mar 2026
Python, FastAPI, Flask, LangChain, LlamaIndex, MCP, Pinecone, ChromaDB, Docker, PostgreSQL, Redis, n8n
- Developed production-ready Generative AI applications using LangChain, LlamaIndex, and vector databases for enterprise knowledge retrieval.
- Adopted MCP (Model Context Protocol) for standardized tool calling across AI agents and deployed the resulting services to production using Docker.
- Fine-tuned open-source LLMs using LoRA and QLoRA, improving model performance for domain-specific AI applications.
- Built real-time AI applications integrating WebSockets, Speech-to-Text, Text-to-Speech, and voice-based conversational AI.
- Automated enterprise workflows using n8n, Zapier, and Zoho, integrating REST APIs, document processing, and business automation services.
- Developed AI-powered web scraping, structured data extraction, backend APIs, and deployment pipelines using Docker, PostgreSQL, Redis, and FastAPI.
Software Developer— Code Insects Pvt. Ltd.
Jan 2024 – Mar 2024
- Built and enhanced scalable web applications following modular and maintainable development practices.
- Implemented robust backend services and API integrations to ensure efficient application functionality and stability.
- Worked closely with development and business teams to translate requirements into reliable technical solutions.
- Performed application testing, troubleshooting, deployment, and maintenance to support stable production environments.
Software Developer— Shine Dezign Pvt. Ltd.
Jun 2023 – Sep 2023
- Developed and maintained scalable web applications with a focus on clean backend architecture.
- Collaborated with cross-functional teams to deliver high-quality software solutions aligned with business requirements.
- Optimized application workflows, resulting in improved system efficiency and user experience.
Project Experience
Healthcare AI – Voice-Enabled Appointment Booking Agent
AI & Backend Engineer
LangGraph, LangChain, FastAPI, MCP, Twilio Voice, OpenAI, Deepgram, ElevenLabs, PostgreSQL, Redis, Docker, WebSockets
- Designed a LangGraph-based multi-agent architecture with intelligent routing, conversation memory, and workflow orchestration for a voice-enabled healthcare scheduling platform.
- Integrated Twilio Voice, Deepgram, and ElevenLabs to enable real-time AI-powered voice conversations with Speech-to-Text and Text-to-Speech capabilities.
- Implemented MCP-based tool calling so agents could securely invoke calendar, EHR, and notification tools through a standardized interface.
- Built scalable backend APIs with FastAPI for calendar scheduling, EHR integration, conversation management, and automated SMS/email reminders.
- Delivered and deployed an end-to-end AI appointment management solution using Docker, reducing manual scheduling effort for healthcare staff.
Enterprise Agentic AI Platform
AI & Backend Engineer
LangGraph, LangChain, OpenAI, FastAPI, Flask, MCP, Pinecone, Qdrant, FAISS, PostgreSQL, Redis, Docker, Python
- Built a production-grade multi-agent platform with planning, reasoning, memory, and human-in-the-loop capabilities for complex business automation.
- Implemented MCP (Model Context Protocol) tool calling to give agents standardized, dynamic access to internal and external tools and enterprise APIs.
- Developed RAG pipelines integrating Pinecone, Qdrant, FAISS, SQL databases, and enterprise REST APIs.
- Implemented concurrent processing using Asyncio, ThreadPoolExecutor, and ProcessPoolExecutor, significantly reducing end-to-end pipeline latency.
- Containerized and deployed the platform using Docker, ensuring scalability and production readiness.
Deepfake Audio Detection System
Machine Learning Engineer
Wav2Vec2, PyTorch, Hugging Face Transformers, Scikit-learn, Python
- Architected an end-to-end deepfake audio detection system, owning the pipeline from data curation and preprocessing through model training, evaluation, and deployment.
- Engineered a transformer-based feature extraction layer using Wav2Vec2 embeddings, benchmarking multiple classical and deep learning classifiers (SVM, Logistic Regression) to identify the optimal architecture for production.
- Drove model performance from baseline to 93.86% classification accuracy in distinguishing real vs. AI-generated speech through systematic feature engineering and hyperparameter tuning.
- Designed and shipped a low-latency, production-grade inference API for real-time audio authenticity detection, with monitoring in place to track model drift and accuracy over time.
Object Classification System
Machine Learning Engineer
PyTorch, OpenCV, ResNet, FastAPI, Docker, Python
- Led the design and development of a real-time computer vision system for industrial image classification, from architecture selection through production rollout.
- Fine-tuned and optimized ResNet-based deep learning models in PyTorch, improving inference accuracy and throughput for high-volume industrial workloads.
- Architected and deployed containerized FastAPI inference services on Docker, building in monitoring, logging, and performance tracking to ensure production reliability.
- Established a reusable model deployment pattern that reduced turnaround time for rolling out subsequent computer vision use cases.
AI Workflow Automation Platform
AI & Backend Engineer
n8n, FastAPI, OpenAI, PostgreSQL, Redis, Python
- Developed AI-powered automation workflows using n8n, FastAPI, and OpenAI models to streamline business processes and reduce manual operations.
- Built intelligent document processing pipelines for extracting, transforming, and structuring information from PDFs and unstructured documents.
- Integrated Gmail, Google Sheets, REST APIs, webhooks, and third-party services to build reliable end-to-end automation workflows.
- Implemented multiprocessing and asynchronous processing techniques to improve workflow throughput and reduce overall processing time.
- Designed scalable AI automation solutions with modular workflows, robust error handling, and efficient data processing for production use cases.
Education
Bachelor of Technology (B.Tech) – Computer Science & Engineering — CGPA: 8.49 / 10.0
Swami Vivekanand Group of Institutes of Engineering & Technology, Mohali, Punjab