Artificial Intelligence Engineer

Rekhta Foundation; Dadri; On-site; INR 1,800,000 - 2,800,000; Full time

Job summary

Rekhta Foundation is seeking an AI Engineer to transform vast cultural archives into intelligent consumer tools and scalable B2B pipelines. You will design and deploy advanced AI systems, including training and fine-tuning open-source models, and build robust RAG pipelines.

You will bridge linguistic data with scalable software engineering for multilingual discovery across Indic languages. The role emphasizes collaboration with language experts and internal teams to automate workflows and align

Qualifications

• Experience building production-grade AI systems and Retrieval-Augmented Generation (RAG) pipelines.
• Experience training and fine-tuning open-source models (LLMs, Vision, Audio).
• Experience integrating with commercial APIs (OpenAI, Gemini) and building structured data outputs.

Responsibilities

• Design production-grade RAG frameworks and embedding-based architectures for semantic search and cross-lingual discovery across Indic languages.
• Train, fine-tune, and evaluate open-source models using LoRA/QLoRA for domain-specific tasks.
• Architect interactions with OpenAI, Gemini APIs, including advanced prompt engineering and function calling.

Skills

Python programming

NLP

Multilingual data handling

Tools

Hugging Face

PyTorch

LangChain

LlamaIndex

Pandas

NumPy

FastAPI

Flask

Qdrant

Milvus

Chroma

pgvector

Docker

Git

API Design

Job description

Role Overview

We are seeking an AI Engineer to transform our massive cultural archives into intelligent consumer tools and scalable B2B pipelines. You will design and deploy advanced AI systems, ranging from training and fine-tuning open-source models to building robust Retrieval-Augmented Generation (RAG) pipelines and leveraging proprietary APIs. You will bridge the gap between core linguistic data and scalable software engineering.

Key Responsibilities

• Core AI & RAG Architecture: Design production-grade RAG frameworks and embedding-based architectures to power semantic search, cross-lingual discovery, and contextual comprehension tools across Indic languages.
• Model Training & Fine-Tuning: Evaluate, train, and fine-tune open-source models (LLMs, Vision, and Audio) using techniques like LoRA/QLoRA for highly domain-specific tasks.
• Proprietary API Integration: Architect advanced interactions with commercial APIs (OpenAI, Gemini), implementing sophisticated prompt engineering, function calling, and strictly structured data outputs.
• Product & Pipeline Automation: Build consumer-facing tools (like automated composition assistants and writing suites) and optimize high-volume enterprise digitization factories (OCR cleaning, layout analysis).
• Cross-Functional Collaboration: Partner with language experts and internal teams to automate workflows and align models with complex cultural and literary nuances.

Technical Requirements

• Programming: Production-grade command of Python and standard data frameworks (Pandas, NumPy, FastAPI/Flask).
• AI/NLP Orchestration: Hands-on experience with Hugging Face (transformers, peft), PyTorch, and frameworks like LangChain or LlamaIndex.
• Embeddings & Vector DBs: Strong understanding of embedding spaces and vector databases (e.g., Qdrant, Milvus, Chroma, pgvector) for building accurate semantic pipelines.
• Software Engineering: Solid grasp of API design, microservices, asynchronous Python, Git workflows, and Docker containerization.

Soft Skills & Culture Fit

• Product Mindset: Focus on the ultimate end-user experience rather than just optimizing standalone test metrics.
• Problem Solver: Excited by the challenge of handling unstructured, messy, and historical multilingual data.
• Cultural Alignment: A deep appreciation for digital humanities and cultural preservation (knowing the specific scripts or languages natively helps, but is not mandatory).