Volunteer for a startup providing accessible mental healthcare solutions across Africa ⭐

CogniX leverages innovative technology to provide accessible, personalized mental healthcare solutions across Africa, empowering individuals and communities to improve their well-being.

CogniX addresses the critical social challenge of limited access to mental healthcare across Africa. With a significant shortage of mental health professionals and widespread stigma surrounding mental health, millions of people in underserved communities lack the support they need. Our platform bridges this gap by providing accessible, affordable, and culturally sensitive mental health solutions. By empowering individuals and communities to prioritize mental well-being, we aim to reduce stigma, improve mental health outcomes, and foster healthier, more resilient communities across Africa. Additionally, our approach promotes environmental sustainability by reducing the need for physical infrastructure in delivering mental health services.

CogniX offers a transformative solution to Africa’s mental healthcare crisis through our comprehensive mobile platform, "Beyond Self." The app connects users with licensed mental health professionals via teletherapy, offering accessible and affordable care regardless of geographic location. Our AI-powered recommendation system matches users with the best-fit therapists based on their needs, while features like mental health tracking, community support, and educational content foster ongoing wellness. Additionally, Beyond Self includes e-pharmacy integration, crisis support, and the use of emerging technologies like VR/AR for immersive therapy experiences, ensuring a holistic and scalable approach to mental health in Africa.

Role (Volunteer, unpaid): AI/LLM Engineer

Role Description: We are seeking a dedicated Voice AI / LLM Engineer to design, deploy, and optimize advanced Large Language Model (LLM)/Agentic-AI systems that power real-time, personalized, and empathetic mental health support. You will lead the development of a production-grade FastAPI backend integrated with Hugging Face-hosted models, retrieval systems, and custom prompt pipelines, bringing meaningful, secure, and impactful AI assistance to users across Africa.

🔧 Key Responsibilities:

• Design and maintain a FastAPI backend that serves LLM responses and connects seamlessly with:
• Hugging Face (for model loading, LoRA adapters, embeddings)
• Vector databases (e.g., Milvus, Chroma).
• Embedding models (e.g., sentence-transformers/all-mpnet-base-v2).
• Custom prompt logic (system prompts, assessments, empathetic templates).
• Fine-tune and optimize LLMs (e.g., LLaMA 3.2+) using LoRA (Low-Rank Adaptation) with Hugging Face's PEFT ecosystem.
• Build and deploy Retrieval-Augmented Generation (RAG) pipelines that use embedding-based search for context-aware responses.
• Develop and document all backend APIs with OpenAPI/Swagger, ensuring clarity and collaboration readiness.
• Implement authentication (JWT), rate limiting, and containerized deployment (Docker) for a secure, scalable API system.
• Lead testing and evaluation of LLM responses, including conversation memory, assessment delivery, and reformulation handling.
• Maintain well-documented, testable, and production-ready code across the LLM stack.

💻 Technologies & Skills:

• Python (advanced proficiency)
• Hugging Face ecosystem: transformers, peft, datasets, accelerate
• Hugging Face Hub integration (private/public repo access)
• FastAPI for backend development and OpenAPI docs
• LoRA / PEFT for parameter-efficient fine-tuning
• Vector databases (Milvus, Chroma, etc.) and semantic search
• LangChain for RAG orchestration (optional but useful)
• Embeddings with sentence-transformers models
• Docker, Git, Jenkins (CI/CD)

Bonus: experience with Streamlit, Gradio, or frontend tech (React/Flutter)

🌟 Key Qualities:

• Strong foundation in LLM architecture, token streaming, and prompt engineering.
• Passion for ethical AI in mental health and social impact sectors.
• Able to take full ownership of model-serving pipelines and backend interfaces.
• Clear communicator and effective collaborator with cross-functional teams.
• Detail-oriented, documentation-driven, and committed to quality.

✅ Benefits:

• Real-world impact: Contribute to accessible, AI-powered mental health support in underserved regions.
• Hands-on experience with Hugging Face, RAG, LoRA fine-tuning, secure API design, and real deployment workflows.
• Peer-Mentorship and recognition: Work with enthusiastic AI/ML team members and earn letters of recommendation and public recognition for your contributions.
• Flexible remote work: Collaborate asynchronously across time zones, ~10+ hours/week.
• Career development: Exposure to full-stack LLM development and potential consideration for future roles.

📬 Application Instructions:

• Please submit your: Updated resume or LinkedIn profile
• Links to GitHub projects or notebooks involving LLMs, Hugging Face, prompt engineering, or backend API work
• (Optional) Brief note highlighting your interest in mental health + AI

Time Commitment: Volunteer 10+ hours per week for 6+ months remotely 💻

If you want to make change, apply to volunteer with CogniX!