About me (Registered since 03/10/2025)

Innovative and results-driven AI Solutions Engineer with over 4 years of experience in designing and deploying end-to-end machine learning solutions across diverse industries, including legal tech, customer support, healthcare, and contract intelligence. Adept at leveraging cutting-edge Generative AI, NLP, LLMs, RAG, and Explainable AI techniques to build scalable, production-grade AI systems.

Proven track record in developing AI-powered chatbots, call center agents, and document processing pipelines using modern frameworks such as LangChain, FastAPI, Flask, and Pydantic. Skilled in integrating VoIP, OCR, and speech synthesis tools (e.g., Twilio, ElevenLabs) to enable multimodal interactions. Experienced in fine-tuning transformer models (Llama, BERT, DistilBERT), implementing automated ML pipelines, and building performance monitoring systems with Dash, Plotly, and XAI.

Recognized for delivering high-impact Proof of Concepts and production-ready solutions that optimize operational efficiency, ensure regulatory compliance, and enhance user experience. Strong understanding of vector databases, microservices, prompt engineering, and cloud-based deployment. Passionate about building intelligent systems that bridge the gap between data and decision-making.

Skills

Tech Skills

Portfolio

Education

Key Skills and Competencies

Programming Languages: Python with OOPs concept, NoSQL, Shell Scripting, Java, C, C++. • Skills: Machine Learning, Data Science, Predictive Modeling, Data Visualization Boosting Bagging Algorithms, Dimensionality Reduction, PCA, TSNE, Clustering Techniques, Generative AI, Conversational AI, Agentic AI, Text to Speech, Speech to Text, HMM, ASR, Neural Networks (ANN, CNN, LSTM, Transformers), NLP, Computer Vision, BERT, T5, GPT, Transfer Learning, CUDA, MPS, Data Analytics, Statistics, AWS EC2, NoSQL, REST API, VoIP, RAG, LLM Fine-tuning (LoRA, QLoRA), VOIP, GIT, ROUGE, BERTScore, Docker, Linux. • Tools: Scikit-learn, TensorFlow, Keras, PyTorch, HuggingFace, Sentence Transformer, LangChain, Pandas, Numpy, Plotly, Dash, Streamlit, Flask, Twilio, PyMongo, Uvicorn, Seaborn, Matplotlib, ElevenLabs, Unsloth, Ollama, ChromaDB, Pinecone.

Work Experience

  • June 2025 - Present
    Thomson Reuters International Pvt. Ltd, India

    AI Solutions Engineer

    • Developed automated XML conversion pipeline for jurisdictional legal documents using Generative AI and NLP techniques, enhancing document processing efficiency.
    • POC on Netscan software to extract the strike-through in the documents using Claude 4 Sonet, OCR and NLP techniques.

  • April 2025 - June 2025
    Starmarks Software Pvt. Ltd, India

    Senior Software Engineer AI/ML

    • Designed a Question Answering AI Chatbot to provide specific addresses, hospital names, and entity meanings using Python, LoRA, QLoRA fine-tuning, Llama 3.2, Hybrid RAG, Unsloth, Vector Database.

  • June 2024 - June 2025
    Capital Numbers Infotech Limited, India

    Senior Software Engineer AI/ML

    • Developed AI-driven call center and customer support agents for inbound and outbound calls, handling use cases like hotel bookings and restaurant reservations using Python, LLM, RAG for Knowledge Based, VoIP (Twilio), ElevenLabs, AI tools, NLP, Async FastAPI.
    • Developed automated prompt generation for all AI agentic infrastructure use cases based on the provided knowledge base using Python, Langchain, Pydantic, GPT 3.5.
    • Implemented advanced analytics to monitor AI agent and user sentiment, with per-tenant pro-rata tracking for actionable insights and performance optimization using Plotly, Dash, XAI techniques.

  • November 2021 - April 2024
    Zycus Infotech Pvt. Ltd, India

    Senior Machine Learning Engineer

    • Built an automated pipeline for training, testing, and deploying AI models for iContract with a single trigger using Python, ML models, transfer learning, microservices, DistilBERT embeddings, TF-IDF, Flask.
    • Developed a performance monitoring tool to track data/model drifts and computational efficiency using Python, Explainable AI, Dash, Plotly, Knowledge Graph, Statistical methods.
    • Fine-tuned NER models and text classifiers for contracts, focusing on specific clauses and entities using Python, Spacy, BERT, ML Statistical Models, Neural Networks.
    • Delivered a custom entity microservice for automating prompt engineering and feature collection using Python, GPT-4, NLP techniques, Flask.

  • July 2021 - September 2021
    VEGG, India

    Machine Learning Developer

    • Built Hot word/Wakeup word model using ASR techniques – MFCC Embeddings and Deep Artificial Neural Network.

Languages

English
Professional
Bengali
Advance
Hindi
Advance