🚀 Data Engineer
💡 Strong foundation in Data Science, Gen AI, building scalable data & AI systems
📍 Hyderabad, India

I’m an Data Engineer at Accordion India with ~2 years of hands-on experience working across data engineering, analytics, and Generative AI.

I enjoy building end-to-end systems — from ingesting and transforming large-scale data to applying ML & LLM-based intelligence and delivering insights through dashboards and APIs.

I’ve worked on production-grade ETL pipelines, BI systems, and GenAI-powered applications involving structured, unstructured, and multi-modal data.

• Python, SQL (T-SQL, PL/SQL)

• ETL / ELT Pipelines
• DBT, Snowflake
• SQL Optimization, Data Modeling
• Git, CI/CD

• Azure (Fundamentals)
• Azure Data Factory (Basics)

• Tableau
• Power BI (Basics)
• Streamlit

• Exploratory Data Analysis (EDA)
• Feature Engineering
• Supervised ML (Scikit-learn)
• Generative AI (LLMs, RAG Systems)
• LangChain

Multi-modal AI-powered retail analytics platform

• Integrated 2+ GB of structured & unstructured data, including Google Reviews, CCTV images/footage, and 1-year transactional store data
• Applied LLMs (GPT-4o mini) for cleanliness scoring, counter utilization insights, and AI-generated feedback
• Used YOLOv8 for footfall & people detection
• Built an interactive Streamlit dashboard with a FastAPI backend
• Implemented a LangChain-powered chatbot for querying store performance

🔗 Tech: LangChain, GPT-4o mini, YOLOv8, FastAPI, Streamlit

End-to-end data & ML pipeline for predicting product dimensions

• Processed ~2GB real-time product data (structured + unstructured)
• Performed data cleaning, feature engineering, and supervised ML modeling
• Built a robust pipeline ensuring consistent handling of missing and noisy data

🔗 Tech: Python, Scikit-learn, Random Forest

Interactive WhatsApp chat analytics application

• Generated 10+ insights on communication patterns for personal & group chats
• Built visualizations for activity heatmaps, emoji usage, URL/file sharing, and sentiment analysis

🔗 Tech: Python, Streamlit, Plotly

Data-driven Olympics analytics dashboard

• Visualized historical insights across all Summer Olympics (till 2016)
• Implemented country-wise, athlete-level, sport-wise, and gender-based analysis

🔗 Tech: Pandas, NumPy, Streamlit, Plotly

• 💼 LinkedIn: https://www.linkedin.com/in/souravchakraborty99
• 🌐 Portfolio: https://iamsourav.pythonanywhere.com/
• 📧 Email: sourav.992001@gmail.com

⭐ If you like my work, feel free to star ⭐ the repositories or reach out for collaboration!