Machine Learning Projects
Detailed showcase of innovative AI and machine learning solutions
Context-Aware Summarization
Context-aware meeting intelligence pipeline that classifies calls, assigns participant roles, routes to specialized workflows, and generates type-specific summaries and insights.
Smart Trackers
Human-in-the-loop few-shot LLM framework that lets users create custom conversation trackers by annotating semantically similar historical examples and applying tracker-specific inference to new meetings.
Customer 360 Dashboard with RAG
Intelligent customer analytics dashboard using RAG technology for comprehensive insights and natural language queries.
Large-Scale Name Matching
Scalable entity matching pipeline for billions of vendor records using fine-tuned Sentence-BERT embeddings, triplet loss, fuzzy matching, and top-K candidate retrieval.
CI Outlines
Meeting transcript outlining service that converts S3-based transcripts into structured hierarchical outlines using text segmentation, lexical signals, embeddings, and LLM-based enrichment.
Address Standardization using IndicBART
Sophisticated address parsing system for Indian addresses supporting multiple languages using fine-tuned IndicBART models.
AI-Based Multimodal Reel Classification
Intelligent content classification system for Instagram reels using multimodal AI approaches, processing both audio transcripts and visual captions.
Lead Prioritization with CatBoost
ML system for financial lead prioritization using gradient boosting algorithms to predict conversion probability.
Tablet Inspection System
Automated visual inspection system for pharmaceutical tablets using deep learning and computer vision technology.
EEG Classification System
Advanced computer vision models for EEG signal classification using CNN architectures to analyze brain activity patterns.
Interested in Collaboration?
I'm always excited to work on innovative machine learning projects and explore new challenges in AI.