Mathew Manoj - Data Science & AI Enthusiast

Hi, I’m Mathew Manoj, an undergraduate at IIT Kharagpur with hands-on experience from internships at Tech Mahindra, Cellstrat, and Ernst & Young. I love building AI solutions like chatbots and predictive models to solve real-world problems. Let’s connect!

Take a look at my work, and feel free to connect if you’re interested in collaborating!

Technical Skills

• Languages: Python, C++
• Libraries/Frameworks: TensorFlow, Pandas, Numpy, Seaborn, Matplotlib, Scikit-Learn, Scipy
• Developer Tools: Visual Studio Code, Google Colab, Jupyter Notebook, MATLAB, XCode
• Cloud Services: AWS (including AWS Bedrock)
• Other Tools: Git

🎓 Education

• M.Tech Dual Degree, Biotechnology

  IIT Kharagpur (Expected 2026)

• Higher Secondary (XII)

  FIITJEE World School (2021)

• Secondary (X)

  FIITJEE World School (2018)

Work Experience

Amazon

Applied Scientist Intern Banglore (May ‘25 - Oct ‘25)

• Fine-tuned a 1.5B LLM for API vs. natural-language response prediction on edge devices.
• Used distillation to help the model decide when to run queries locally or offload them.
• Reached high accuracy in both API invocation and decision routing for production use.

Prodoc AI

Data Science Intern Banglore (Dec ‘24)

• Trained DeBERTa for end-of-conversation detection with strong accuracy and precision.
• Built a VAD module to filter speech from noise, improving conversation flow and timing.
• Deployed the model on Hugging Face and integrated it into healthcare apps, boosting user interaction

Tech Mahindra Maker’s Lab

AI Research Intern
Pune (Jun ‘24 - Sep ‘24)

• Did research on deploying LLM models onto edge devices like raspberry pi.
• Contributing to Project Indus, building LLMs for Indian languages.
• Engaging in foundational AI research for scalable AI systems.

Cellstrat Inc

Data Science Intern
Bangalore (Jan ‘24 - Present)

• Enhanced a Knowledge Mining chatbot with AWS services.
• Developed a chatbot using the Amazon Titan model for an AWS workshop.
• Evaluated over 50 different voices and 30-40 prompts for PhoneCallGPT.

Ernst and Young (EY) LLP

Data Analytics Intern
Bangalore (Jun ‘23 - Jul ‘23)

• Developed a 3-layer LSTM for sales prediction with an MAE of 0.0919.
• Implemented LeakyReLU activation and MSE loss for high prediction accuracy.

Follicular Stage Identification Project

Computer Vision Research Intern
IIT Kharagpur (May ‘23 - Present)

• Utilized 2D fluorescent images to categorize ovarian follicle stages.
• Analyzed images to measure cellular thickness and identify growth phases.

🛠️ Projects

Emotion-Led Data Analysis | Self Project (Aug ‘24)

• Conducted exploratory data analysis (EDA) on candidate profiles to understand emotion-led patterns in recruitment.
• Visualized trends in emotions, sentiments, and personality traits using Python libraries like Matplotlib and Seaborn.

• Developed insights to enhance data-driven recruitment strategies.

• GitHub Repository

PubMed Paper Scraper | Self Project (July ‘24)

• A web app built with Streamlit to search and retrieve academic papers from PubMed based on user-defined topics.
• Displays key information like paper title, journal, publication date, and abstract with links to PubMed.

• Features error handling, rate limiting, and search capabilities for relevant papers.

• PubMed Paper Scraper Website
• GitHub Repository

---

PDF to HTML Converter | Self Project (August ‘24)

• A web app that converts LinkedIn PDF resumes into professional HTML format using OpenAI’s API.
• Built using Streamlit for the interface and PyPDF2 for extracting text from PDFs.

• Generates well-formatted HTML resumes based on extracted content.

• PDF to HTML Converter Website
• GitHub Repository

---

Chatbot Development Using AWS | Intern Project (Jan ‘24)

• Built a chatbot with AWS Bedrock and AWS Lambda.
• Achieved successful output of up to 20 tokens, encountering challenges with higher token limits.
• Implemented a web adapter layer for streaming responses.

More Projects

🏆 Competitions & Conferences

Meta Hackathon (Oct ‘24)

• Created a WhatsApp-based conversational agent using Meta LLaMA models (LLama - 2.1B) to analyze customer satisfaction and identify product needs
• Used Python, LangChain, MongoDB, and ChromaDB to do the context retrieval
• GitHub Repository

Unit Liability Prediction Model | Inter Hall Data Analytics Competition (Feb ‘23 - Mar ‘23)

• Developed a model predicting manufacturing unit safety with 97% accuracy, made it to 1st place in accuracy.

Construction of Fintech Website | Inter Hall Opensoft Competition (Mar ‘23)

• Developed a SaaS platform for financial services using React.js, Node.js, and MongoDB.

🏅 Positions of Responsibility

Academic Head | Kharagpur Game Theory Society | IIT Kharagpur (Aug ‘23 - Apr ‘24)

• Organized workshops on game theory concepts and coordinated the flagship event Stratathon.

⚽ Extracurricular Activities

• Football: Active member of the bronze-winning football team of Vidyasagar Hall of Residence.
• Swimming: Participant in events with the Technology Aquatics Society at IIT Kharagpur.
• Cultural: Organized Pongal celebrations for IIT Kharagpur Tamil Sangam, attracting over 200 attendees.

🔗 Let’s Connect!

• GitHub
• LinkedIn
• Twitter

Feel free to explore my projects, and don’t hesitate to reach out if you’d like to collaborate or have any questions!