AI Engineer – LLM & RAG SystemsCompany: Monarch Digital Assets - Email your resume to [email protected]

Experience: 0–2 Years

Location: New Delhi

Employment: Full-Time

RoleMonarch Digital Assets is building an LLM-based system for long-form digital publishing. The platform ingests multiple books on a subject, extracts and structures their information, identifies relationships across sources, and uses that knowledge to generate recent chapter-based manuscripts with source traceability and measurable information coverage.

We are looking for a AI Engineer to work directly on this pipeline. This is a hands-on engineering role covering document processing, RAG, knowledge graphs, LLM orchestration and evaluation. You will work closely with the project lead initially and progressively take ownership of core components of the system.

What You Will Build

- Build ingestion pipelines for PDF, EPUB and DOCX documents, including parsing, chunking, metadata extraction and document hierarchy preservation.
- Develop RAG pipelines covering embeddings, indexing, semantic/hybrid retrieval, reranking and context construction.
- Extract structured facts, concepts, arguments, examples, entities and relationships from long-form documents using LLMs and schema-based outputs.
- Work with vector databases and graph databases to support cross-document retrieval and GraphRAG-style knowledge synthesis.
- Build multi-stage LLM workflows for extraction → planning → retrieval → generation → verification → revision.
- Develop methods to track source provenance and measure whether generated chapters have omitted, duplicated or distorted source information.
- Implement LLM evaluation pipelines for retrieval quality, factual grounding, information coverage and hallucination detection.
- Build backend services and APIs for model execution,



document processing and asynchronous generation workflows.
- Benchmark alternative models, retrieval strategies, chunking methods and prompts using measurable evaluation criteria rather than manual prompt testing alone.
- Optimise pipelines for accuracy, latency, token usage and inference cost.

Required Skills
- Strong Python programming fundamentals.
- Experience working with LLM APIs such as OpenAI, Anthropic, Gemini or open-source models.
- Understanding of embeddings, vector search and Retrieval-Augmented Generation (RAG).
- Working knowledge of SQL, REST APIs and Git.
- Experience with FastAPI, Flask or similar Python backend frameworks.
- Good understanding of data structures, debugging and software engineering fundamentals.
- Ability to independently investigate technical problems and evaluate competing approaches.

Good to Have
- LangGraph, LlamaIndex, LangChain or similar LLM orchestration frameworks.
- PostgreSQL/pgvector, Qdrant, Pinecone, Weaviate, FAISS or similar vector stores.
- Neo4j, knowledge graphs or GraphRAG.
- PyMuPDF, Docling, Unstructured or other document-processing libraries.
- LLM evaluation tools such as Ragas, DeepEval, TruLens or custom evaluation frameworks.
- Docker, cloud deployment, async Python, queues and background workers.
- Basic understanding of transformers, NLP and PyTorch.

Candidate ProfileWe are open to recent graduates and engineers with limited commercial experience if they have built technically credible AI projects.

A strong GitHub project involving RAG, document intelligence, knowledge graphs, LLM evaluation or multi-step LLM workflows will carry more weight than certificates alone.

This role is intended for someone who wants early ownership of a real LLM product. As the platform develops, you will have the opportunity to own major parts of its AI architecture and build a portfolio around production-grade RAG, GraphRAG, long-context document processing and LLM evaluation.

📌 AI Engineer - LLM & RAG Systems (Delhi)
🏢 Monarch Digital Assets
📍 Delhi