AI Systems Professional

I build AI systems that translate complex data into actionable decision intelligence, combining machine learning, multimodal data processing, and agentic LLM-based architectures. My work focuses on end-to-end system development, from problem formulation and modeling to deployment and real-world usage, across domains including finance, healthcare, and industrial AI.

What I Work On

  • Agentic AI & Decision Systems:
    Designing multi-agent systems that decompose complex user queries into structured workflows for reasoning, analysis, and decision support across real-world applications
  • LLM-based Automation & Code Generation:
    Building domain-specific systems for natural language-driven analytics (text-to-SQL, text-to-R), enabling non-technical users to create and execute data workflows
  • Data Transformation & ML Pipelines:
    Developing systems to standardize and process noisy, heterogeneous datasets into structured formats, improving downstream modeling and analytics efficiency
  • Scalable AI Platforms & Applied ML Systems:
    End-to-end development of AI systems combining machine learning, multimodal data processing, and deployment pipelines for real-world decision-support applications

Selected Work

  • Multi-Agent Financial Analysis Assistant: [GitHub]
    Built an agentic AI system that decomposes natural language queries into specialized sub-tasks (technical, fundamental, and sentiment analysis), enabling structured investment research and evaluation of real-world market scenarios
  • Domain-Specific Code Generation System:
    Developed an LLM-based system enabling text-to-SQL and text-to-R workflow generation over structured datasets, reducing manual query and script development effort and lowering the barrier for non-technical users
  • LLM-Driven Data Transformation Pipeline:
    Designed a system to standardize noisy, heterogeneous datasets into structured formats, reducing data preparation effort by up to 40% and enabling more efficient downstream analytics
  • AI Systems for Anomaly Detection & Condition Assessment:
    Built multimodal machine learning systems for automated inspection and state estimation, contributing to reusable AI platforms and decision-support systems across multiple industrial applications

Technical Skills

  • AI / Machine Learning:
    Deep Learning, Transformers, LLMs, VLMs, Generative AI, Computer Vision, Multimodal Learning, Supervised & Unsupervised Learning
  • GenAI Systems & Agentic AI:
    LangChain, LangGraph, MCP, A2A, RAG systems, CrewAI, Flowise, Prompt/Context Engineering, Vector Embeddings, Ollama
  • AI Systems & Deployment:
    AWS (Bedrock, Lambda, SageMaker, Boto3), Docker, FastAPI, MLflow, ONNX, Serverless Architectures, Hugging-Face
  • Programming & Data Engineering:
    Python, PyTorch, OpenCV, NumPy, Pandas, Scikit-learn, Open3D, R, TypeScript, Git

Education

  • PhD in Computer Science (graduated 2021): SMU, Singapore.
    Dissertation: Vision-based Analytics for Improved AI-driven IoT Applications [ Link ]
  • M.Tech in Mobile & Ubiquitous Computing (graduated 2014): IIIT Delhi, India
  • B.Tech in Computer Science (graduated 2012): GJ University of Science & Technology, India

Professional Certificates