🚀 Soumik Banerjee | AI Engineer & Data Science Enthusiast
🌟 About Me
I’m an AI Engineer and Data Science enthusiast dedicated to solving complex challenges with innovative, data-driven solutions. Currently pursuing my M.Sc. in Data Science at VIT, I continuously question the status quo to build and optimize systems that not only perform but also adapt to future demands. My journey spans from developing cutting-edge deep research agents to deploying AI-powered applications that automate critical processes, always with a focus on efficiency and scalability.
🎓 Education
-
M.Sc. in Data Science
Vellore Institute of Technology, Vellore (08/2023 - 05/2025)
CGPA: 8.87/10 -
B.Sc. in IT (Data Science)
Maulana Abul Kalam Azad University of Technology, West Bengal (08/2020 - 07/2023)
CGPA: 9.72/10
💼 Professional Experience
-
Analytics & Artificial Intelligence Intern @ Pinak Idea Lab
May 2024 - Present- Spearheading the development of an open-source deep research agent using DeepSeek R1 and Tavily, capable of generating comprehensive research reports from 50+ URLs within minutes.
- Automating file query and retrieval (RAG) processes through GPT-4o-mini, Gemini 2.0, Zoho, Supabase, and n8n.
- Engineered a Streamlit application for semantic clustering and search volume analysis—saving 61 hours and reducing API costs by 98%.
-
Data Science Intern @ Think-Again-Lab, Kolkata
Dec 2022 - Mar 2023- Leveraged predictive analytics with R and Tableau to forecast trends and perform correlation analysis on extensive datasets.
- Led projects to identify emerging programming trends and optimize strategies through deep data insights.
🚀 Projects
-
Generalized Medical Recommendation System using Deep Research Agent (Jan 2025 - Present)
- Developed an autonomous AI research agent that curates and synthesizes information from a self-collected 150-person medical dataset to offer timely, reliable healthcare suggestions.
- Explore the project
-
Cold Email Generator using Llama 3.1 (Oct 2024 - Nov 2024)
- Engineered a tool with Llama 3.1, LangChain, ChromaDB, and Streamlit to automate the creation of personalized outreach emails, challenging conventional marketing methods.
- Explore the project
-
PyroAlert: AI-Powered Fire Detection (Sep 2024 - Dec 2024)
- Designed an AI-driven fire detection system using YOLO for real-time monitoring, achieving a remarkable 0.93 mAP in performance.
- Explore the project
-
State-wise Business Comparison and Forecasting (Jan 2024 - Jul 2024)
- Built a dynamic Streamlit application to visualize and forecast business activities across 28 Indian states using ARIMA, driving smarter decision-making.
- Explore the project
-
Speech Emotion Recognition using LSTM (Dec 2022 - Apr 2023)
- Pioneered an LSTM-based model in TensorFlow to analyze speech emotions, achieving a 98% accuracy on the Toronto Emotion Speech Set, questioning and refining emotion detection methodologies.
- Explore the project
🚀 Skills & Tools
- Programming Languages: Python, SQL
- Frameworks & Libraries: TensorFlow, Transformers, GPT, Gemini, Llama, DeepSeek, BERT
- Tools & Platforms: n8n, LangGraph, Streamlit, MySQL, Supabase, Postgres
- Expertise: Data Automation, Generative AI, Prompt Engineering, Retrieval Augmented Generation (RAG), Machine Learning, Deep Learning, NLP, Data Visualization, Predictive Analytics
- Soft Skills: Collaboration, Data-Driven Decision-Making, Communication, Leadership