Popai Health-Applied AI Engineer

Description:

Popai Health is a fast-growing health tech startup transforming the way care teams engage with

patients. We build AI tools that analyze phone conversations between care managers and patients,

surfacing insights, automating documentation, and helping close care gaps—so teams can focus on

what matters most: delivering better care.

What you'll do:

• Algorithm Development: Build on top of cutting-edge LLMs by crafting custom infrastructure and logic for validation, pre-/post-processing, and model refinement—delivering tailored, production-grade capabilities.
• Benchmark and Compare Models: Establish frameworks for models comparison on unique healthcare data in multiple languages, focusing on cost-efficiency, inference speed, and accuracy.
• Open-source LLM Exploration: Investigate the feasibility of running LLMs locally with limited hardware resources, including model quantization, pruning, and efficiency optimizations.
• STT Model Analysis: Analyze and compare Speech-to-Text (STT) models, including fine-tuning state-of-the-art models for domain-specific accuracy.
• Efficiency Optimization: Propose and implement methods to reduce inference costs and improve performance without compromising accuracy.
• Collaboration: Work closely with data engineers and software developers to integrate optimized models into production workflows.
• Documentation: Maintain clear and detailed documentation for model evaluations, fine-tuning processes, and efficiency benchmarks.

• 1+ year of experience in developing, deploying, and managing LLMs in production environments, including building supporting infrastructure, testing, and evaluation.
• Proficiency in Python and ML frameworks (PyTorch, LangChain, Hugging Face).
• Experience with Speech-to-Text models, especially fine-tuning and evaluation.
• Knowledge of on-premises deploymenttechniques, including quantization and model compression for limited hardware.
• Experience with cloud infrastructure (AWS, GCP) and cost optimization strategies for LLMs
• Ability to work independently, taking ownership of tasks and driving them to completion.