Senior AI Lead | AI Solution Architect

Senior AI Lead | AI Solution Architect

Principal Systems Engineer with 18+ years of experience designing, building, and modernizing large-scale enterprise platforms.

Over the last year, I have upskilled in AI and Generative AI, focusing on designing and delivering production-grade AI solutions, including Agentic AI systems, AI agents, and RAG-based architectures.

I specialize in applying AI to real enterprise problems—modernizing legacy systems, automating workflows, and building scalable, cloud-native, and secure AI platforms. My work bridges strong engineering fundamentals with modern AI architecture and product thinking.

• Agentic AI systems: multi-agent workflows, orchestration, tool use
• GenAI product architecture: RAG patterns, evaluation, guardrails, observability
• AI product ownership: turning business problems into scalable AI products

• Built a serverless AI Digital Twin application with a conversational interface, using React, FastAPI on AWS Lambda, Amazon Bedrock (LLM), S3, CloudFront, and Terraform for automated infrastructure provisioning. Repo: https://github.com/madhusudhanan-jayaram/twin

• Vehicle Damage Detection (Deep Learning + FastAPI + HTML UI)

• Built a deep-learning–based computer vision system to detect and classify vehicle damages from images. Exposed the model via FastAPI with a simple HTML UI for real-time inference and demo usage. Repo: https://github.com/madhusudhanan-jayaram/deeplearning_vehicle_damage_detection

• WhatsApp Analyzer (NLP + analytics + web app structure)

• Developed an NLP-driven analytics engine to extract insights from WhatsApp chat data. Performed sentiment analysis, activity trends, and message statistics with a structured web app interface. Repo: https://github.com/madhusudhanan-jayaram/whatsapp_analyser

• Stock Prediction (Time Series)

• Implemented time-series forecasting models to predict stock price movements using historical data. Applied feature engineering and model evaluation techniques to analyze trends and volatility. Repo: https://github.com/madhusudhanan-jayaram/stock_prediction_time_series

• Movie Recommendation System

• Built a recommendation engine using collaborative and content-based filtering techniques. Generated personalized movie suggestions based on user preferences and similarity metrics. Repo: https://github.com/madhusudhanan-jayaram/movie_recommendation_system

• Kidney Disease Prediction (Machine Learning + Flask Web App) Built an end-to-end ML application for chronic kidney disease prediction with a Flask web interface deployed on Google Cloud App Engine, leveraging Vertex AI for model training, experimentation, and scalable inference within a production-ready cloud architecture.

  Repo: https://github.com/madhusudhanan-jayaram/2.Machine-Learning

• Portfolio: https://madhusudhanan.online
• GitHub: https://github.com/madhusudhanan-jayaram

• AI/ML:

• GenAI:

• Backend:

• Frontend:

• Platform & CI/CD:

• API Gateway:

• Data & Messaging:

• Vector DB:

• Observability & Security:

🚀

I’m always interested in practical use-cases where agents + good architecture + product thinking create real business value.

👁️