AIML/Gen AI Developer

VMC Soft Technologies, Inc; Bengaluru; On-site; INR 3,500,000 - 7,000,000; Full time

Job summary

VMC Soft Technologies, Inc. is seeking a Generative AI Developer to build an advanced RAG pipeline with agentic and multimodal capabilities for processing complex contract documents (PDFs with text, tables and images).

You will work with open-source LLMs, design systems that combine vision and language models for deeper contract understanding. Ideal candidates have 6+ years in backend/ML, 1+ year in GenAI/RAG, strong Python skills, and hands-on experience with LangChain, LlamaIndex and modern

Qualifications

• 6+ years in backend/ML
• 1+ year in GenAI/RAG
• Strong Python skills; experience with LangChain, LlamaIndex, Haystack
• PDF parsing experience (text, tables, images)
• Multimodal processing experience with GPT-4V/LLaVA/Donut
• Hands-on with vector DBs, FastAPI, Redis/Postgres
• Understanding of legal/contract structure and NLP challenges

Responsibilities

• Build end-to-end RAG pipelines: parse chunk embed retrieve generate
• Develop agentic workflows using AgenticX, LangGraph, or custom orchestration
• Parse contract PDFs using layout-aware and OCR-based parsers (Unstructured, PyMuPDF, PDFPlumber)
• Enable multimodal processing (text + images/tables) using models like GPT-4V, LLaVA, Donut
• Integrate open-source models (Mistral, LLaMA3, Ollama) and vector databases (Milvus, FAISS, Qdrant)
• Build modular APIs (FastAPI) for Q&A, summarization, and classification across documents
• Optimize retrieval/generation for long-context and multi-document inputs

Skills

Backend/ML development

GenAI/RAG systems

Python

LangChain

LlamaIndex

PDF parsing

FastAPI

Redis

Postgres

NLP challenges

Tools

Unstructured

PyMuPDF

PDFPlumber

Milvus

FAISS

Qdrant

GPT-4V

LLaVA

Ollama

LangGraph

AgenticX

Job description

About The Role

We’re hiring a Generative AI Developer to build an advanced RAG pipeline with agentic and multimodal capabilities for processing complex contract documents (PDFs with text, tables, and images). You'll work with open-source LLMs, build agent workflows with AgenticX/LangGraph, and design systems that combine vision and language models for deeper contract understanding.

Experience: 6+ years

Skills: Backend/ML development, 1+ year with GenAI/RAG systems

Location: Bangalore, HYD, Kochi, Delhi NCR, Bhubaneswar and Pune

Responsibilities

• Build end-to-end RAG pipelines: parse chunk embed retrieve generate
• Develop agentic workflows using AgenticX, LangGraph, or custom orchestration
• Parse contract PDFs using layout-aware and OCR-based parsers (Unstructured, PyMuPDF, PDFPlumber)
• Enable multimodal processing (text + images/tables) using models like GPT-4V, LLaVA, Donut
• Integrate open-source models (Mistral, LLaMA3, Ollama) and vector databases (Milvus, FAISS, Qdrant)
• Build modular APIs (FastAPI) for Q&A, summarization, and classification across documents
• Optimize retrieval/generation for long-context and multi-document inputs

Required Skills

• 6+ years in backend/ML; 1+ year in GenAI, RAG, or agentic architectures
• Strong Python skills; experience with LangChain, LlamaIndex, Haystack
• Deep experience in PDF parsing (text, tables, images) and chunking strategies
• Familiarity with multimodal models (e.g., GPT-4V, LLaVA, or equivalents)
• Hands-on with vector DBs, FastAPI, Redis/Postgres
• Understanding of legal/contract structure and NLP challenges

Bonus Skills

• Experience with LangGraph, AgenticX, or custom multi-agent systems
• Familiarity with multimodal fusion techniques and prompt engineering
• Knowledge of document layout analysis and long-form reasoning
• Prior work in legal tech, compliance, or enterprise contract analytics