05/06/2026 updated
DG
100 % available
AI Engineer | RAG & GenAI Systems | Data Engineering & NLP
Bengaluru, India
Worldwide
B.Tech in Computer Science and Engineering, National Institute of Technology Silchar (CGPA: 8.18)Generative AI & RAG Systems
Expertise in designing and building end-to-end Retrieval-Augmented Generation pipelines including hybrid retrieval, semantic search, chunking strategies, contextual compression, re-ranking, citation-based QA, and evaluation using metrics such as Precision@k, Recall@k, MRR, NDCG, faithfulness, and answer relevance.
Python & Programming
Strong foundations in Python, SQL, C++, FastAPI, REST APIs, OOP, and DSA, with practical application in building scalable AI systems, data pipelines, and backend services.
Data Engineering
Hands-on experience building scalable data ingestion pipelines using PySpark, Pandas, NumPy, Databricks, Delta Lake, ETL/ELT workflows, CDC loads, schema validation, and data quality checks for structured enterprise data sources.
LLMs & NLP
Practical knowledge of OpenAI APIs, Hugging Face Transformers, Sentence Transformers, LangChain, LlamaIndex, transformer-based embeddings, text classification, and summarization.
Vector Search & Information Retrieval
Experience with FAISS, pgvector, Chroma, Pinecone basics, BM25, metadata filtering, top-k retrieval, indexing strategies, cosine similarity, and evaluation metrics including Precision, Recall, MRR, and NDCG.
Machine Learning & Deep Learning
Proficiency with PyTorch, TensorFlow, Scikit-learn, XGBoost, LightGBM, regression, classification, model evaluation, cross-validation, and hyperparameter tuning.
Deployment & DevOps Tools
Familiarity with Docker, Git, GitHub Actions, Streamlit, Next.js basics, Kubernetes basics, AWS/GCP basics, logging, monitoring, and unit testing for deploying AI applications.
Knowledge Graphs
Integration of Neo4j knowledge graphs to store entities, relationships, document topics, authors, methods, datasets, and citations for graph-enhanced retrieval in RAG systems.
Expertise in designing and building end-to-end Retrieval-Augmented Generation pipelines including hybrid retrieval, semantic search, chunking strategies, contextual compression, re-ranking, citation-based QA, and evaluation using metrics such as Precision@k, Recall@k, MRR, NDCG, faithfulness, and answer relevance.
Python & Programming
Strong foundations in Python, SQL, C++, FastAPI, REST APIs, OOP, and DSA, with practical application in building scalable AI systems, data pipelines, and backend services.
Data Engineering
Hands-on experience building scalable data ingestion pipelines using PySpark, Pandas, NumPy, Databricks, Delta Lake, ETL/ELT workflows, CDC loads, schema validation, and data quality checks for structured enterprise data sources.
LLMs & NLP
Practical knowledge of OpenAI APIs, Hugging Face Transformers, Sentence Transformers, LangChain, LlamaIndex, transformer-based embeddings, text classification, and summarization.
Vector Search & Information Retrieval
Experience with FAISS, pgvector, Chroma, Pinecone basics, BM25, metadata filtering, top-k retrieval, indexing strategies, cosine similarity, and evaluation metrics including Precision, Recall, MRR, and NDCG.
Machine Learning & Deep Learning
Proficiency with PyTorch, TensorFlow, Scikit-learn, XGBoost, LightGBM, regression, classification, model evaluation, cross-validation, and hyperparameter tuning.
Deployment & DevOps Tools
Familiarity with Docker, Git, GitHub Actions, Streamlit, Next.js basics, Kubernetes basics, AWS/GCP basics, logging, monitoring, and unit testing for deploying AI applications.
Knowledge Graphs
Integration of Neo4j knowledge graphs to store entities, relationships, document topics, authors, methods, datasets, and citations for graph-enhanced retrieval in RAG systems.
Languages
EnglishFluent
Project history
Designed an end-to-end RAG system for research papers, technical PDFs, and enterprise documents with ingestion, parsing, chunking, embedding, indexing, retrieval, generation, and evaluation modules. Built a hybrid retrieval pipeline combining dense vector search using Sentence Transformers, keyword-based BM25 retrieval, metadata filtering, and cross-encoder re-ranking. Integrated Neo4j knowledge graph and implemented contextual compression and citation-based answering. Containerized the backend and vector database using Docker and exposed APIs through FastAPI.
Built a production-style AI support agent that answers customer queries from PDFs, FAQs, product manuals, tickets, and policy documents using RAG and tool-calling workflows. Designed an agentic workflow with planner, retriever, response generator, escalation classifier, and tool execution modules. Added multimodal document processing support, feedback logging, evaluation scripts, and safety and governance checks. Deployed the application with Docker, FastAPI backend, vector database service, and a Next.js/Streamlit frontend.
Developed a semantic search system using transformer-based embeddings to match resumes with job descriptions. Built a FAISS index for efficient similarity search and contextual ranking of candidate profiles. Improved resume-job matching beyond keyword search using embedding-based semantic similarity.