Sachin Kumar
Machine Learning & AI Enthusiast
Aspiring AI/ML Engineer with expertise in Machine Learning, Deep Learning, and Generative AI. Passionate about solving real-world problems using AI-driven products.
Contact Me
Technical Expertise
Programming
Python, C++
AI/ML
TensorFlow, PyTorch, Scikit-learn, Deep Learning
NLP/GenAI
Hugging Face, LangChain, RAG, FAISS
Data & DB
NumPy, Pandas, SQL, MySQL, MongoDB
Deployment Tools
  • AWS Cloud Services
  • FastAPI & Streamlit
  • Git/GitHub
  • Linux Systems
Visualization
  • Matplotlib
  • Seaborn
  • Plotly
Education & Background
1
IIIT Guwahati
B.Tech, Computer Science & Engineering
Expected 2026 — CGPA: 6.55/10
Relevant Coursework: Data Structures, OOP, Machine Learning, Artificial Intelligence
2
S. S. Public School
XII (CBSE) — 86.40%
X (CBSE) — CGPA 9.6/10
Smart Loan Recovery System
Project Overview
Designed an end-to-end loan recovery prediction system analyzing borrower income, loan size, and repayment history. Achieved 85% accuracy using Random Forest Classifier.
  • K-Means clustering for borrower segmentation
  • Automated pipeline reducing manual effort by 40%
  • Plotly dashboards for business insights
Tech Stack
Python, Pandas, Scikit-learn, Plotly
85%
Accuracy
Random Forest model performance
40%
Efficiency Gain
Reduction in manual analysis
3
Risk Groups
Low, medium, high segments
Fashion Recommendation System
CNN-Based Engine
Leveraged VGG16 feature extraction for outfit similarity across 97+ product images.
Deep Features
Applied cosine similarity for recommendation ranking with automated preprocessing pipeline.
Top-5 Recommendations
Generated personalized outfit suggestions, improving user experience and product discovery.
Tech Stack: Python, TensorFlow/Keras, NumPy, SciPy, Matplotlib
Context-Aware Chatbot (RAG)
Data Ingestion
Trafilatura & PyPDF extract content from web pages and documents
Vector Storage
FAISS stores semantic embeddings for fast vector search
LLM Integration
Hugging Face Transformers power intelligent responses
Real-Time Chat
Streamlit deployment enables user interaction
Key Features
  • Retrieval-Augmented Generation (RAG)
  • Relevance gate reducing irrelevant responses by 20%
  • Scalable pipeline for enterprise integration
Technologies
Python, Hugging Face, FAISS, PyTorch, Streamlit
Project Impact & Results
85%
Model Accuracy
Loan recovery prediction system
40%
Efficiency Boost
Automated analysis pipeline
20%
Response Quality
Improved chatbot relevance
Smart Loan Recovery
Segmented borrowers into risk groups, enabling targeted recovery strategies with visual dashboards.
Fashion Recommendations
Processed 97+ images with VGG16, delivering personalized top-5 outfit suggestions using deep learning.
RAG Chatbot
Built context-aware AI assistant with FAISS vector search and LLM integration for accurate Q&A.
Achievements & Recognition
Problem Solving Excellence
Solved 300+ Data Structures & Algorithms problems across LeetCode, GeeksforGeeks, and Coding Ninjas, strengthening analytical and coding skills.
Tata Motors Hackathon
Designed a Forward Collision Detection System prototype documentation, applying Machine Learning in safety-critical automotive systems.

Continuous Learning: Completed Data Science & AI Bootcamp by Krish Naik on Udemy (May 2025), staying current with industry best practices.
Skills & Competencies
Machine Learning
  • Supervised Learning
  • Unsupervised Learning
  • Deep Learning
  • Model Optimization
Generative AI
  • Large Language Models
  • RAG Systems
  • Vector Databases
  • Prompt Engineering
Software Engineering
  • API Development
  • Cloud Deployment
  • Version Control
  • Agile Practices
01
Data Collection & Preprocessing
Extract, clean, and prepare data for analysis
02
Model Development & Training
Build and optimize ML models for performance
03
Deployment & Monitoring
Deploy solutions with FastAPI/Streamlit on AWS
Let's Connect
Ready to Build AI Solutions
Passionate about leveraging AI and Machine Learning to solve real-world problems. Open to opportunities in AI/ML Engineering, Data Science, and Generative AI development.
📧 Email
📱 Phone
+91-9610339676
Made with