Summary
Overview
Work History
Education
Skills
Certification
Accomplishments
Interests
Timeline
Generic
Ayan Debnath

Ayan Debnath

AI/ML Engineer
Kolkata, West Bengal

Summary

Enthusiastic AI/ML Engineer with 5 years of experience in designing and deploying intelligent systems across NLP, Gen AI, and Agentic AI domains. Skilled in transformer-based models, fine-tuning, RAG pipelines, and agentic architectures for enterprise and EdTech solutions. Proven track record of delivering scalable AI applications. Adept at bridging research-driven innovation with real-world deployment. Seeking a challenging position to contribute technical proficiency and strategic thinking in a dynamic and collaborative environment.

Overview

6
6
years of professional experience
1
1
Certification
3
3
Languages

Work History

IT Analyst

Tata Consultancy Services
02.2024 - Current
  • EdTech Tutor using Agentic AI: Designed a multi-agent AI tutor system that autonomously generates course modules, quizzes, and personalized explanations based on subject, grade, and syllabus. Integrated long-term memory with a vector database (Qdrant/Solr) to track student progress and tailor future lessons.
  • Enterprise Chatbot: Built an agentic AI enterprise chatbot capable of orchestrating multi-tool calls across HRMS, Learning, etc platforms and knowledge base systems for unified query resolution. Designed a multi-agent architecture where different agents collaborate to decompose queries, call APIs, and validate responses. Implemented retrieval-augmented memory (RAG) using a vector database for context-aware, cross-departmental knowledge retrieval. Implemented short-term conversation storage (Cassandra/Redis/MongoDB) for seamless multi-turn interactions and long-term memory (Qdrant/Milvus) to preserve context across conversations, personalize responses, and reduce repeated queries.
  • Hybrid Search Engine with Vector DB: Implemented a hybrid search engine (semantic + keyword + filtering + reranking) using vector databases (Qdrant/Solr) to improve course and job recommendations in TCS iON Learning Hub and Job Portal ensuring both accuracy and coverage.
  • Question Duplication Checker: Developed a question duplication detection system to identify plagiarized or duplicate questions across online sources and internal question bank databases. Implemented Google Search API + web scraping pipelines to fetch potential duplicates from the web in real time. Added fuzzy string matching & lexical overlap metrics for shallow checks, combined with semantic clustering for deep context-aware detection.

Systems Engineer

Tata Consultancy Services
08.2020 - 01.2024
  • Domain-specific Question Authoring: Developed an automated question bank authoring system using a QLoRA-finetuned Mistral model on TCS iON exam papers, enabling domain-specific question generation. Integrated a Multimodal RAG pipeline with a vector database to ground question generation in curriculum documents or textbooks (pdf/docx/image), ensuring factual alignment and coverage. Implemented Bloom’s Taxonomy–based reasoning to scale difficulty levels. Added flexibility to switch engines by supporting Azure OpenAI GPT models alongside Mistral, giving institutions a choice between open-source fine-tuned and enterprise-grade managed models.
  • Multilingual Translation: Finetuned a state-of-the-art transformer model on TCS iON’s large-scale question paper dataset to enable domain-aware, context-sensitive translation of academic and exam content. Optimized translations for exam-specific terminology, outperforming generic machine translation baselines in accuracy and contextual fidelity. Incorporated flexibility to switch between the finetuned in-house model and Google Translate API for scalability and fallback. Enhabled multi-lingual delivery of question papers, supporting regional language adoption in large-scale assessments.
  • ID card OCR: Developed functionality to detect and extract information from various Govt. ID card images or PDFs using Object Detection and OCR, achieving an accuracy rate of 84%. Deployed as a REST API for seamless integration with enterprise verification workflows (KYC, HR onboarding, access management).

Machine Learning Intern

Tata Consultancy Services
01.2020 - 04.2020
  • AI-driven Survey Analysis: Applied Bidirectional LSTM with attention mechanism for emotion analysis on survey responses, achieving improved context-sensitive sentiment detection.
  • Implemented keyphrase extraction and semantic clustering (using embeddings + clustering algorithms) to group responses and reveal major themes.
  • Combined deep learning and unsupervised techniques to generate actionable insights and highlight key themes from unstructured text data.

Education

Bachelor of Technology - Information Technology

Institute of Engineering and Management
01.2020

Higher Secondary - Science

Ramakrishna Mission Rahara
01.2016

Skills

Machine Learning

undefined

Certification

IBM AI Engineering (07/2022 - Present) - Professional certification offered by IBM through Coursera on building blocks of Machine Learning and Deep Learning

Accomplishments

  • Innovation Superstar Award from TCS iOT IDE (03/2023 - Present) Awarded by TCS iOT IDE Connected assets Team for winning machine learning based hackathon on steam turbine health prediction and providing business context on the problem statement.
  • Special Initiative Award from TCS iON (02/2023 - Present) Awarded by TCS iON for outstanding contribution.
  • On The Spot Award from TCS iON (07/2021) Awarded in appreciation of contribution and commitment towards team.
  • Departmental Topper in Institute of Engineering and Management (2017 - 2020) Awarded as Topper in Information Technology department.

Interests

Driving, Travel, Sports

Timeline

IT Analyst

Tata Consultancy Services
02.2024 - Current

Systems Engineer

Tata Consultancy Services
08.2020 - 01.2024

Machine Learning Intern

Tata Consultancy Services
01.2020 - 04.2020

Higher Secondary - Science

Ramakrishna Mission Rahara

Bachelor of Technology - Information Technology

Institute of Engineering and Management
Ayan DebnathAI/ML Engineer