AI Engineer Consultant – Health Data Platform

About the role

PATH is looking for an AI Engineer Consultant to enhance SnapiForm, an AI‑powered platform that digitises health‑worker paper forms via mobile images. The role will focus on building robust computer‑vision and vision‑language pipelines that can accurately read handwritten text, extract complex tables and parse multilingual data in low‑resource settings.

Key responsibilities

• Design, develop and optimise AI pipelines for document understanding, targeting handwriting recognition, table extraction and multilingual parsing.
• Research, benchmark and fine‑tune state‑of‑the‑art Vision‑Language Models (e.g., Qwen‑VL) and OCR models on domain‑specific datasets.
• Apply advanced techniques such as LoRA/QLoRA and DeepSpeed to improve accuracy on noisy mobile images.
• Architect and deploy production‑grade inference services using vLLM or similar engines, managing continuous batching, KV‑cache and quantisation for cost‑effective throughput.
• Plan and provision cloud (Linode) and on‑prem GPU infrastructure, evaluating options like NVIDIA Blackwell nodes to support up to 10 million forms per month.
• Develop image‑preprocessing strategies (chunking, tiling) to handle high‑resolution documents and large tables.
• Prepare the system for future offline/edge processing, expanded language support and integration beyond DHIS2.
• Travel to PATH sites as required.

Required profile

• Strong background in AI, computer vision and large‑scale model deployment.
• Experience with Vision‑Language Models and OCR technologies.
• Proven ability to optimise inference pipelines for low‑cost, high‑throughput environments.
• Willingness to work in a global nonprofit focused on health equity.

Required skills

• Computer vision
• Vision‑Language Models (e.g., Qwen‑VL)
• OCR models
• LoRA / QLoRA
• DeepSpeed
• vLLM or similar inference engine
• Continuous batching, KV‑cache management, quantisation
• Cloud platforms (Linode) and GPU hardware (NVIDIA Blackwell)

What we offer

• Opportunity to impact health data systems in low‑resource settings.
• Collaboration with a multidisciplinary global nonprofit.
• Professional growth in cutting‑edge AI research and deployment.