AHAMED ZAHEE KHAN M

AHAMED ZAHEE KHAN M

Generative AI Engineer
India
Tamil, English
Full-time On-site

نبذة عني

Generative AI Engineer with 3+ years of hands-on experience designing, fine-tuning, and deploying enterprise-grade Large Language Model (LLM) systems, Retrieval-Augmented Generation (RAG) pipelines, and NLP automation to…

الخبرة

AI Solutions Engineer – Generative AI

Fintech Software Solutions · Chennai
Jun 2023 - حتى الآن · 3 سنوات 4 أشهر

<p>Architected and deployed Retrieval-Augmented Generation (RAG) pipelines using Pinecone and Chroma vector databases, enabling semantic document search across 10,000+ enterprise documents with >90% retrieval accuracy.</p><p> Designed and integrated LLM workflows using LangChain and GPT-4 for domain-specific enterprise Q&A, reducing average document lookup time by 60%.</p><p> Fine-tuned LLaMA 2 (7B) models using LoRA and QLoRA parameter-efficient fine-tuning on internal knowledge bases, reducing hallucination rate by ~35% and improving factual accuracy for internal Q&A tasks.</p><p> Implemented LLM evaluation frameworks (RAGAS) and guardrail mechanisms to enforce output reliability, safety, and accuracy in production Q&A systems.</p><p> Engineered NLP pipelines for automated document extraction, named entity recognition, classification, and summarization, processing 500+ business documents per day with 92% accuracy.</p><p> Automated extraction and classification of enterprise business documents using GPT-4 function calling and custom prompt chains, eliminating 15+ hours of manual review weekly.</p><p> Built production-grade conversational AI chatbots and LLM agents with persistent memory, multi-turn context, and domain-specific reasoning using LangChain and FastAPI.</p><p> Built and deployed FastAPI inference services for LLM model serving on AWS EC2, with S3-backed document storage and Lambda-based event triggers.</p><p> Created PySpark-based ETL workflows on AWS, achieving a 40% improvement in pipeline performance and reducing data processing latency from 4 hours to under 2.5 hours.</p><p> Collaborated with engineering and cloud teams to design scalable AI architectures; documented best practices and conducted internal Generative AI training sessions for 20+ team members.</p>

المشاريع

RAG-Based Enterprise Knowledge Assistant

Vestas
المدة : 07-Dec-2023 - 03-Jul-2024

<p> ● Built a GenAI-powered document Q&A assistant using LangChain, GPT-4, and Pinecone for enterprise-wide contextual knowledge retrieval, with chunking strategies, OpenAI Ada-002 embedding pipelines, and conversational retrieval chains with chat history.</p><p> ● Reduced employee document lookup time by 60% and decreased support tickets by 30% within 3 months of deployment. </p>

الشهادات

Master of Computer Application

University of Madras, India · Chennai , Tamil Nadu , India · 2023

المهارات

GPT-4 GPT-3.5 LLaMA 2 LLaMA 3 Mistral RAG Prompt Engineering LangChain LangGraph LangChain Agents AI Agents Agentic Workflows Embeddings Fine-Tuning LoRA QLoRA RLHF Model Evaluation Guardrails RAGAS Pinecone Chroma FAISS Semantic Search Embedding Pipelines Natural Language Processing (NLP) Hugging Face Transformers PyTorch TensorFlow Classification Regression Feature Engineering Named Entity Recognition (NER) Text Summarization Python SQL PySpark Git ETL Pipelines Apache Airflow Data Quality Distributed Computing FastAPI REST APIs Docker MLflow AWS Azure CI/CD MLOps
الإبلاغ عن هذا الملف الشخصي؟