Role Overview

Key Responsibilities

- Design and implement end-to-end RAG pipelines integrating vector databases (FAISS, Pinecone, Weaviate, ChromaDB) with LLMs for accurate, grounded responses.
- Build and optimize document ingestion pipelines including chunking strategies, embedding generation, and metadata filtering.
- Implement conversational agents with memory management (buffer, summary, vector store memory) and multi-step reasoning.
- Architect stateful multi-agent systems using LangGraph including cyclic reasoning loops, conditional edges, and human-in-the-loop workflows.
- Design and deploy OCR pipelines for processing scanned documents, PDFs, invoices, and forms using engines like Tesseract, PaddleOCR, and EasyOCR.
- Implement document layout analysis to extract tables, headers, key-value pairs, and structured fields from unstructured documents.
- Integrate OCR outputs with downstream NLP and LLM pipelines for intelligent document processing (IDP).
- Monitor agent performance, debug hallucination issues, and implement guardrails for production deployments.
- Collaborate with product and data engineering teams to translate business requirements into scalable AI solutions.
- Document technical designs, architecture decisions, and system performance benchmarks.

Required Skills & Qualifications

- Strong hands-on experience with RAG frameworks (LlamaIndex, LangChain, Haystack) and vector databases.
- Proficiency with LangChain v0.2+ including LCEL, agents, tools, and structured output generation.
- Experience building stateful agent graphs with LangGraph including nodes, edges, and state schemas.
- Hands-on experience with OCR engines (Tesseract,



PaddleOCR, EasyOCR) and cloud document AI APIs (Google Vision, AWS Textract, Azure Form Recognizer).
- Proficiency in Python with experience in async programming, REST APIs, and microservices.
- Familiarity with LLM providers such as OpenAI, Anthropic Claude, Gemini, Cohere, or Mistral.
- Experience with embedding models (OpenAI Ada, BGE, SentenceTransformers) and similarity search.
- Knowledge of image pre-processing and computer vision libraries (OpenCV, Pillow, scikit-image).
- Familiarity with cloud platforms (AWS, GCP, Azure) for model hosting, storage, and deployment.
- Understanding of prompt engineering, output parsers, and LLM evaluation techniques.

Preferred / Good to Have

- Experience with multi-modal RAG (text + image) and hybrid search (BM25 + dense retrieval).
- Knowledge of agentic patterns such as ReAct, Plan-and-Execute, Reflexion, and supervisor agents.
- Experience with document AI models such as LayoutLM, Donut, or TrOCR.
- Exposure to Intelligent Document Processing (IDP) platforms and workflow automation tools.
- Knowledge of multi-lingual OCR and Indic script recognition.
- Contributions to open-source AI/ML or NLP projects.

Educational Qualification

- B.E. / B.Tech / M.Tech in Computer Science, Information Technology, or a related field.
- Equivalent practical experience with a strong portfolio of AI/ML projects will also be considered.

Pay: ₹1,000,000.00 - ₹1,500,000.00 per year

Benefits :

- Cell phone reimbursement
- Versatile schedule
- Health insurance
- Internet reimbursement
- Leave encashment
- Life insurance
- Provident Fund

Work Location: In person

📌 Ai/ml Engineer (Pune)
🏢 Royal Sports E Fitness
📍 Pune