Home

Param Madan - AI/ML Engineer

madanparam18@gmail.com

Location: Boston, Massachusetts, USA

Relocation: Open

Visa: F1

Resume file: Param_Madan_1785785353850.pdf Please check the file(s) for viruses. Files are checked manually and then made available for download.

Param Madan Boston, MA, USA | +1 617-259-8214 | madanparam18@gmail.com | LinkedIn | GitHub Summary AI/ML Engineer with 3+ years of experience designing, building, and deploying scalable machine learning systems and intelligent applications across high-growth tech, HR-tech, and enterprise environments. Hands-on expertise with Python, Apache Spark, PyTorch, TensorFlow, FastAPI, AWS, and Kubernetes, with a strong foundation in Generative AI, Large Language Models (LLMs), Retrieval-Augmented Generation (RAG), MLOps, Apache Kafka, MLflow, and Databricks. Proven track record delivering production-ready AI solutions that process millions of daily transactions, reduce operational costs, and accelerate decision-making for cross-functional teams. Experienced in end-to-end ML life-cycle management from feature engineering and model training to CI/CD pipelines, experiment tracking, and real-time model monitoring at scale. Adept at collaborating with engineering, product, and data science stakeholders to translate complex business requirements into cloud-native, enterprise-grade AI/ML platforms. Experience AI/ML Engineer Mar 2026 Present Uber USA Architected real-time demand forecasting pipelines using Python, PySpark, and Apache Spark across 100+ metropolitan zones, enabling accurate driver supply planning during peak demand. Engineered pricing models using Apache Kafka, XGBoost, and Feature Stores, deploying 12+ production models for fare estimates and surge recommendations on high-volume transactions. Optimized online inference services using PyTorch, Docker, and Kubernetes to process 800K+ prediction re-quests daily with sub-second response times for rider and driver applications. Orchestrated ML workflows using Apache Airflow, MLflow, and Amazon SageMaker with 6 cross-functional stake-holders to streamline model validation and production release cycles. Championed generative AI initiatives leveraging LangChain, Vector Databases, and LLM frameworks with 2 engi-neering teams to deliver support assistants and knowledge retrieval workflows. AI/ML Consultant Jul 2024 Dec 2024 Massachusetts Institute of Technology - Spinout (AutonomUS) USA Architected a battlefield surgical robotics system to automate wound detection across 14 trauma scenarios using real-time ultrasound imaging and CNNs via cross-functional clinician workshops. Optimized deep learning U-Net models within PyTorch and CUDA to eliminate 3 probabilistic risk conditions by anchoring code changes in thorough peer reviews. Orchestrated production pipelines tracking model drift across 25,000 validation ultrasound frames by translating technical risks directly into concrete milestones for executive project stakeholders. Overhauled the end-to-end architecture planning and deployment of AI models to hit a 0.8-second processing latency threshold by unifying engineering teams on workflows. Championed system deployment to satisfy 4 military medical standards by building compliance verification frame-works rooted in technical ingenuity and cross-team peer alignment. Software Developer Jun 2021 Jul 2023 Dell Technologies India Developed backend modules using Python, FastAPI, and PostgreSQL managing 25K+ enterprise devices across 8 operational workflows for internal asset management applications. Engineered data processing components using Pandas and SQL to collect, clean, and transform 500K+ device events for reporting dashboards and predictive analytics. Integrated prediction services with Scikit-learn and XGBoost into backend APIs enabling hardware health assessments across 5 infrastructure monitoring systems. Automated build and deployment using Docker and Jenkins for 15+ Python services across development, testing, and production environments. Implemented ML integration workflows by validating and deploying 6+ predictive models into enterprise support platforms, coordinating API contracts with data science teams. Projects LLM-Powered Multi-Agent RAG System | LangChain, Pinecone, FastAPI, AWS ECS 2024 Built a multi-agent RAG pipeline with dynamic document retrieval over 50K+ knowledge base entries, achieving 87% response accuracy using semantic chunking and reranking strategies. Deployed as a containerized FastAPI microservice on AWS ECS with auto-scaling, serving 5K+ queries/day with under 200ms latency. Real-Time Fraud Detection Platform | Kafka, Spark Streaming, XGBoost, Kubernetes 2023 Engineered a streaming fraud detection system processing 10K+ transactions/second using Apache Kafka and Spark Streaming with XGBoost ensemble models achieving 96% precision. Deployed on Kubernetes with MLflow experiment tracking, reducing false positive rate by 23% over baseline models through iterative feature engineering. End-to-End MLOps Platform | MLflow, Airflow, Docker, AWS SageMaker, Terraform 2024 Designed a fully automated MLOps platform orchestrating model training, versioning, and deployment using MLflow and Apache Airflow, cutting release cycle time from 3 days to under 4 hours. Provisioned scalable cloud infrastructure on AWS SageMaker with Terraform, enabling one-click model promo-tion across dev, staging, and production environments with drift monitoring alerts. Intelligent HR Chatbot with Semantic Search | Hugging Face, FAISS, FastAPI, React.js 2024 Built a domain-specific conversational AI assistant using Hugging Face sentence-transformers and FAISS vector indexing over 200K+ HR policy documents, reducing ticket volume by 35%. Exposed inference via a FastAPI REST layer with JWT authentication, integrated into a React.js portal used by 8K+ employees across 4 business units. Distributed Time-Series Forecasting Engine | PySpark, Prophet, Databricks, Delta Lake 2023 Developed a distributed forecasting engine on Databricks using PySpark and Prophet to generate 30-day demand forecasts across 500+ SKUs, improving inventory accuracy by 19%. Stored feature snapshots and model outputs in Delta Lake with time-travel support, enabling reproducible backtests and audit-ready compliance reporting. Technical Skills Programming Languages: Python, Java, R, SQL, TypeScript, JavaScript, Go, Bash AI & Machine Learning: Scikit-learn, TensorFlow, PyTorch, XGBoost, Pandas, NumPy, Hugging Face, NLP, Feature Engineering, Predictive Modeling, Deep Learning, Model Evaluation Generative AI & LLM Engineering: LLMs, RAG, LangChain, Prompt Engineering, AI Agents, Semantic Search, Embedding Models, Vector Databases (Pinecone, FAISS, ChromaDB), LLMOps, VLLM Data Engineering: Apache Spark, PySpark, Apache Kafka, Apache Airflow, Databricks, Hadoop, Dask, ETL/ELT Pipelines, Batch & Real-Time Data Processing, Data Modeling Cloud, DevOps & MLOps: AWS (EC2, S3, Lambda, ECS, EKS, SageMaker, Glue, Redshift), GCP (BigQuery, Cloud Run), Azure (Azure ML, Synapse, AKS), Docker, Kubernetes, Terraform, MLflow, CI/CD Backend & Software Engineering: FastAPI, Flask, Django, Spring Boot, Node.js, REST APIs, GraphQL, Microser-vices, Event-Driven Architecture, OAuth 2.0, OOP, System Design, Agile Databases & Storage: PostgreSQL, MySQL, MongoDB, Redis, Snowflake, Neo4j, Oracle, Azure Cosmos DB Developer Tools: Git, GitHub Actions, Jenkins, Linux, Tableau, Power BI, Alteryx Education Northeastern University Boston, MA, USA Master of Science in Engineering Sep 2023 Dec 2025 Keywords: continuous integration continuous deployment artificial intelligence machine learning javascript business intelligence sthree database rlang golang Massachusetts