AI Engineer (Modeling) - Expression of Interest

ThinkMD is a high-growth, AI-native health and data company transforming global healthcare through scalable, clinically validated technology. Our mission is to defeat global healthcare inequity by enabling mass distribution of AI-driven clinical decision-support solutions.

Already conducting 22,000+ clinical assessments per week across our partner countries including Zambia, Kenya, Sierra Leone and Cambodia, ThinkMD is rapidly scaling across the developing and low resource countries — providing powerful health intelligence that optimizes patient & clinician experiences, informs national-level health decision making, and drives commercial value.

Backed by world-class venture partners and global health leaders, ThinkMD is not just creating impact—we are commercializing the future of AI-powered healthcare in low-resource settings. Through strategic partnerships with governments, not for profits, private healthcare providers, pharma, and telecom companies, we are positioning ourselves as a market leader in digital health solutions for the Global South and beyond.

We're scaling globally to impact over 1 billion lives by 2030. To support this next phase of innovation and growth, we're seeking a talented AI Engineer (Modeling) to strengthen our technical capabilities which underpin our products and organizational capabilities.

Ready to help us impact a billion lives by 2030? Submit your Expression of Interest today!

The Opportunity
As an AI Engineer (Modeling) at ThinkMD, you will focus on post-training and optimization of large language models (LLMs) and other AI systems, including fine-tuning, RLHF, and model distillation. You will also develop and apply advanced prompting strategies, RAG, and other approaches to improve performance and repeatability. Collaborating with cross-functional teams, you will drive model performance to support our mission of equitable healthcare delivery.

What You'll Do

Model Post-Training & Optimization
- Conduct fine-tuning, RLHF, and model distillation using techniques like Supervised Fine-Tuning (SFT), Quantized Low-Rank Adaptation (QLORA), Proximal Policy Optimization (PPO), and Direct Preference Optimization (DPO)
- Apply compression and optimization techniques to improve model efficiency without sacrificing performance
- Implement RLHF pipelines using libraries like trlX and Hugging Face TRL to align models with human preferences

Retrieval & Prompt Engineering
- Integrate LLMs with external knowledge sources to ensure factual accuracy and up-to-date responses
- Design and implement prompting strategies (e.g., building upon recent advances in Chain of Thought, Tree of Thought, and Chain of Verification etc.) to enhance model reasoning

Evaluation & Responsible AI
- Design and implement evaluation metrics and A/B testing to assess model performance and reliability in production
- Develop strategies to ensure ethical AI use, including bias mitigation, fairness, and transparency in model outputs

Research & Collaboration
- Work with pipeline engineers and data scientists to integrate optimized models into production pipelines
- Stay updated on advancements in AI/ML, experiment with new techniques, and contribute to ThinkMD's AI strategy

What You Bring

Education & Experience
- Master's or PhD in Computer Science, Electrical Engineering, Statistics, or a related field, with formal training in AI/ML, including coursework in deep learning, reinforcement learning, and/or natural language processing
- 2+ years in AI/ML engineering, with a focus on model post-training and optimization
- Proven expertise in fine-tuning and applying RLHF to LLMs (e.g., Llama, DeepSeek, Qwen)
- Experience with prompting and RAG-based strategies to improve model reasoning and reliability

Technical Expertise
- Deep Learning Frameworks: Proficiency in PyTorch and Hugging Face Transformers
- Tools and Frameworks: Familiarity with trlX, TRL, Unsloth, and Weights & Biases (Wandb) for model training and evaluation
- Prompting Techniques: Expertise in advanced prompting strategies (e.g., Chain of Thought, Tree of Thought, Chain of Verification)
- Programming: Advanced proficiency in Python; familiarity with shell scripting is a plus

Professional Qualities
- Strong analytical skills
- Excellent communication
- Ability to collaborate in a fast-paced, team-oriented environment

Preferred Qualifications
- Experience with model distillation and compression for efficient LLM deployment
- Knowledge of best practices for model evaluation, including metrics design and A/B testing
- Familiarity with responsible AI frameworks and techniques for bias mitigation
- Contributions to open-source AI/ML projects or publications in conferences like NeurIPS, ICLR, or ICML

What We Offer
- Opportunity to make a meaningful impact on global healthcare inequity
- Collaborative work environment with a mission-driven team
- Competitive compensation and benefits package
- Professional growth and development opportunities
- Flexible work arrangements

Requirements
- Ability to work flexible hours across global time zones with global teams
- Legal work rights in your country of residence
- Clean background check

Ready to help us impact a billion lives by 2030?
Submit your Expression of Interest today!

ThinkMD is an equal opportunity employer committed to building a diverse and inclusive team. We welcome candidates from all backgrounds who share our passion for using technology to transform healthcare accessibility worldwide.