Abraham Jeevan Roy

Core Expertise

Generative AI

Expertise in Large Language Models (LLMs), prompt engineering, and building AI-powered applications using GPT, Claude, and open-source models. Creating innovative solutions with RAG architectures and fine-tuning techniques.

Machine Learning

Deep understanding of ML algorithms, model training, and deployment. Experience with TensorFlow, PyTorch, scikit-learn, and building end-to-end ML pipelines for production environments.

Data Engineering

Proficient in data preprocessing, feature engineering, and working with large-scale datasets. Expertise in pandas, NumPy, and building efficient data pipelines for ML workflows.

NLP & Search

Advanced natural language processing with transformers, embeddings, and semantic search. Building intelligent search systems using FAISS, vector databases, and hybrid retrieval methods.

Model Deployment

Experience deploying ML models to production using Flask, FastAPI, and cloud platforms. Building scalable APIs and integrating AI solutions into real-world applications.

Performance Optimization

Optimizing model inference, implementing quantization, and improving system performance. Experience with batch processing, caching strategies, and efficient resource utilization.

Featured Projects

RAG-Based Document Q&A System

Built an intelligent document question-answering system using Retrieval-Augmented Generation. Implemented multi-model embedding configurations with FAISS indexing for efficient semantic search.

Python LangChain FAISS Azure OpenAI

Semantic Search Engine

Developed a production-ready semantic search API with cross-encoder reranking and confidence scoring. Optimized for high-performance retrieval across large document collections.

Flask Sentence Transformers Vector DB REST API

AI Content Summarization Tool

Created an automated document summarization pipeline supporting multiple file formats. Implemented batch processing with metadata extraction and customizable summary generation.

NLP Transformers PDF Processing Gradio

Local LLM Deployment Framework

Developed an efficient framework for deploying quantized LLMs locally. Implemented 4-bit quantization and optimized inference for resource-constrained environments.

GGUF Quantization Model Optimization GPU Acceleration

Embedding Configuration Manager

Built a comprehensive UI for managing multiple embedding model configurations. Features include dynamic config generation, FAISS index creation, and persistent storage management.

Gradio Embeddings Azure AI Configuration Management

Hybrid Search System

Implemented a hybrid search combining dense embeddings and sparse retrieval methods. Integrated cross-encoder reranking for improved relevance and accuracy.

Information Retrieval Reranking FAISS Performance Tuning

Experience & Journey

AI/ML Developer

Specializing in Generative AI Solutions

Developing cutting-edge AI solutions with focus on document processing, semantic search, and RAG systems. Built production-ready applications using Azure OpenAI, LangChain, and various embedding models.

Machine Learning Engineer

Building Intelligent Search Systems

Designed and implemented semantic search engines with FAISS indexing, cross-encoder reranking, and hybrid retrieval methods. Optimized model performance and deployed scalable API solutions.

Data Scientist

NLP & Document Processing

Developed NLP pipelines for document analysis, summarization, and information extraction. Worked with transformers, embeddings, and built efficient batch processing systems.

Python Developer

API Development & System Integration

Created RESTful APIs using Flask and FastAPI. Integrated various AI services, managed deployment pipelines, and ensured system reliability and performance.