Hello, I'm
Susanta Baidya
AI/ML Professional | Gen AI & Agentic AI Enthusiast
AI/ML Professional | Gen AI & Agentic AI Enthusiast
Passionate AI/ML engineer with expertise in cutting-edge technologies
AI/ML Intern at DeepMatrix
MSc AI & ML, IIIT Lucknow
Bengaluru, India
I am Susanta Baidya, an AI/ML professional currently working as an AI/ML Intern at DeepMatrix, Bengaluru. I hold an MSc in Artificial Intelligence and Machine Learning from IIIT Lucknow (CGPA: 8.20/10) and a BSc in Statistics.
With hands-on experience spanning Computer Vision, Generative AI, and Agentic AI through industry internships, I am passionate about developing cutting-edge AI solutions including LLM-powered applications, autonomous AI agents, and production-ready ML systems.
My work ranges from YOLO-based object detection systems to sophisticated RAG implementations and natural language to command-line interfaces. I have successfully contributed to systems that improved inference speed by 35% and detection accuracy by 28%.
Industry experience in AI/ML development and real-world problem solving
Academic foundation in AI/ML and statistical sciences
Artificial Intelligence & Machine Learning
Statistics
Indian Space Research Organisation (ISRO)
AIR 218 (out of 3563) - Top 6.12%
171st of 3000+ participants (97/100)
0.81 score in PII Detection challenge
Comprehensive expertise across AI/ML technologies and modern development tools
Comprehensive portfolio of AI/ML projects spanning Computer Vision, Generative AI, and Agentic AI
Advanced computer vision system using YOLOv5 for detecting 23 defect classes, achieving 0.56 mAP@0.5 with Flask-based API and real-time visualization.
Complete RAG pipeline with semantic search using Sentence Transformers and FAISS vector indexing, processing 2,507+ quotes with 100% query success rate.
Advanced agentic AI system with fine-tuned LLM that autonomously converts natural language instructions into precise command-line operations with intelligent planning.
Deep learning-based prediction system using bidirectional LSTM, achieving 69% accuracy with 12-hour lead time for Uttar Pradesh weather forecasting.
Advanced image preprocessing pipeline for dental IOPA X-rays, enhancing diagnostic accuracy while maintaining clinical workflow efficiency.
ML pipeline that automatically classifies customer support tickets by issue type and urgency level with intelligent entity extraction capabilities.
Deep RNN model trained on five years of historical Google stock price data to forecast stock performance over a two-month period with high accuracy.
Python web scraper that extracts comprehensive project information from the Odisha RERA website with automated data processing and analysis.
Implementation of neural network fundamentals using Python and NumPy, covering weight initialization, forward/backward propagation, and gradient descent optimization.
Generative AI application leveraging Google's Gemini LLM that analyzes uploaded images to generate unique creative stories, demonstrating multimodal AI capabilities.
User-centric Streamlit application for forecasting flight ticket prices. Achieved 99% accuracy using Random Forest Regression with comprehensive feature engineering.
Intelligent recommendation system using Streamlit and BERT-based NLP, providing tailored book suggestions from Goodreads dataset based on genre preferences.
Ready to discuss AI/ML opportunities, collaborations, or innovative projects!
Bengaluru, India
+91-8837403472
susantabaidya20133@gmail.com