Senior AI/ML Engineer - Document Intelligence & Generative AI

Job Title : Senior AI/ML Engineer (Document Intelligence & Generative AI)

Location : Chennai - Onsite - Hybrid

Notice Period : Joining within 30 Days only

Job Summary :

We are seeking a Senior AI/ML Engineer with 5-8+ years of experience to lead the development of our next-generation document processing platform. In this role, you will move beyond simple predictive modeling to design end-to-end intelligent systems that "read" and understand complex documents (PDFs, Images, Excel, PPT).

You will architect and deploy high-performance pipelines combining OCR, Document Layout Analysis, Vision-Language Models (VLMs), and Large Language Models (LLMs). You will be responsible not just for model experimentation, but for the full lifecycle-from architecture and quantization to scaling and production deployment.

Key Responsibilities :

1. System Architecture & Orchestration :

- Design end-to-end ML system architectures that orchestrate multiple components : OCR engines, embedding models, Vector DBs, and LLM prompt pipelines.

- Build robust preprocessing pipelines for diverse unstructured data formats (Images, PDFs, PPTs, Excel).

- Design async processing queues and microservices to handle high-concurrency workloads.

2. Advanced Model Implementation :

- OCR & Layout Analysis : Implement and fine-tune pipelines using tools like Tesseract, PaddleOCR, or cloud APIs (AWS Textract, Google Vision). Deploy document layout models like LayoutLMv3, Donut, or LiLT.

- Vision-Language Models : Leverage state-of-the-art VLMs (e.g., CLIP, BLIP-2, LLaVA, Florence-2) for multimodal understanding.

- Generative AI & LLMs : Implement information extraction using OpenAI/Claude APIs or host open-source models (Llama 3, Qwen, Mistral) for private deployments.

3. Performance Optimization & Scaling :

- Optimize ML systems for production scaling, ensuring low latency and high throughput for batch processing.

- Apply model optimization techniques, including distillation and quantization (GGUF, ONNX) for efficient CPU/GPU inference.

- Implement deployment strategies that balance cost and performance across GPU and CPU resources.

4. MLOps & Production Engineering :

- Build CI/CD pipelines for ML (using GitHub Actions, Jenkins) to automate testing and deployment.

- Implement monitoring solutions to track concept drift, latency, and inference costs.

- Develop and expose models via robust REST APIs.

Required Technical Qualifications :

- Experience :

• 5-8+ years of hands-on experience in Machine Learning / Deep Learning.
• At least 2-3 years in a Senior or Lead role actively building production systems.

• Expert proficiency in Python and ML frameworks (PyTorch, TensorFlow).

• OCR : Tesseract, EasyOCR, PaddleOCR, AWS Textract, or Google Vision.
• Layout Models : LayoutLMv3, Donut, Pix2Struct, LiLT, LayoutXLM.
• VLMs : CLIP, BLIP-2, GroundingDINO, LLaVA, Florence-2.

• Experience with LLMs (OpenAI, Gemini, Llama, Mistral).
• Strong knowledge of Prompt Engineering, RAG (Retrieval-Augmented Generation), and Vector Databases.

• Proficiency in API development (FastAPI/Flask).
• Experience with Docker, container orchestration, and async task queues (Celery/Kafka).
• Experience optimizing models (ONNX, GGUF, Quantization).