Senior AI Engineer

Job description

# Senior AI Engineer > 2070Health · Bengaluru, India · Full-time · Posted 2026-04-14 **Workplace:** on_site **Department:** 2070 Health ## Description ### Description **About Contiinex** Contiinex is an AI-first enterprise automation platform for healthcare and insurance, purpose-built to understand unstructured conversations, documents, and workflows, and autonomously execute complex, human-intensive business processes. We build specialised domain-trained Small Language Models (SLMs) and fine-tuned LLM pipelines designed to operate reliably in regulated, high-stakes environments such as US Healthcare Revenue Cycle Management (RCM). Our architecture emphasizes deterministic AI systems combining prompt engineering, model fine-tuning, and agentic orchestration to power real enterprise automation. **Role Overview** We are seeking a Senior AI Engineer with strong expertise in Prompt Engineering, LLM fine tuning, and Small Language Model (SLM) development to design, train, optimize, and deploy domain-specialised language models. A key focus of this role will be engineering high-performance prompts for 8B-class models (such as LLaMA, Mistral, and Qwen) and transitioning these prompts into fine-tuned models for production reliability. You will design prompt architectures, instruction schemas, and evaluation pipelines that ensure models produce accurate, structured, and deterministic outputs suitable for enterprise automation workflows. **Key Responsibilities** ● Design production-grade prompt architectures for 8B-class models. ● Develop structured prompts for enterprise tasks such as classification, extraction, reasoning, and summarization. ● Optimize prompts for accuracy, latency, and cost efficiency. ● Build prompt evaluation frameworks to measure accuracy, hallucination rates, and consistency. ● Design reusable prompt libraries and prompt templates for enterprise workflows. ● Develop prompt-to-model migration strategies converting high-performing prompts into fine-tuned SLMs. ● Design and fine-tune LLMs for domain-specific enterprise tasks. ● Develop Small Language Models (SLMs) optimized for enterprise deployment. ● Build instruction tuning and supervised fine-tuning (SFT) pipelines. ● Design evaluation datasets and automated benchmarking frameworks. ● Implement retrieval augmented generation (RAG) pipelines and tool-augmented workflows. ● Collaborate with speech AI and document AI teams to build multimodal systems. ● Deploy models in private cloud or on-premise environments with strong security controls. **Required Qualifications** **Education** Master’s degree or PhD in Computer Science, AI, Machine Learning, or a related field. **Experience & Technical Skills** ● Strong expertise in Prompt Engineering for 7B–13B models (especially 8B models). ● Experience designing prompts for structured enterprise outputs. ● Experience building prompt evaluation datasets and benchmarking frameworks. ● Ability to convert prompt workflows into fine-tuned models. ● 4–6 years of experience in ML/NLP with 3+ years focused on LLMs or foundation models. ● Hands-on experience fine-tuning open-source models such as LLaMA, Mistral, Falcon, or Qwen. ● Experience with LoRA, QLoRA, adapters, and model distillation techniques. ● Strong understanding of transformers, tokenization, embeddings, and attention mechanisms. ● Strong Python engineering skills and experience with PyTorch. **AI Platform & Infrastructure** ● Experience with GPU-based training and inference. ● Familiarity with Hugging Face, Accelerate, DeepSpeed, and Triton. ● Experience with vector databases and RAG architectures. ● Experience deploying models using Docker, Kubernetes, and cloud platforms. Compliance & Enterprise Readiness ● Experience working in regulated environments. ● Understanding of data privacy, access controls, and AI auditability. ● Ability to design AI guardrails and human-in-the-loop workflows. ## Apply [Apply at 2070Health](https://apply.workable.com/2070health/j/2746FFB908/apply) --- Powered by [Workable](https://www.workable.com)