HacamaTech

HacamaTech V6.0: Bridging 1,400 Years of Traditional Wisdom with Gemini 3’s Spatial Intelligence.

Comment

Inspiration

I discovered the Gemini 3 Hackathon just few days ago, on February 4th. Driven by a deep curiosity about how cutting-edge AI could modernize ancient healing traditions, I embarked on a solo journey. My goal was to see if Gemini 3’s multimodal capabilities could bring surgical precision and data-driven safety to Hijama (cupping) and Leech therapy—practices that have remained largely unchanged for 1,400 years.

What it does

HacamaTech V6.0 acts as a 'Clinical Co-pilot.' It uses Gemini 3’s spatial reasoning to analyze images of a patient’s anatomy and identify precise treatment points based on traditional protocols. Beyond mapping, it integrates a holistic dashboard featuring Hijri calendar optimization for treatment timing and personalized post-therapy nutrition guidance.

  • Holistic Integration: Features dedicated modules for Leech Therapy (Sülük) and Post-Therapy Nutrition.

  • Lunar Intelligence: Automatically calculates Hijri "Gold Days" (17th, 19th, 21st) for optimal treatment timing.

How I built it

The project is powered by Gemini 3 Flash for high-speed image reasoning. I utilized System Instructions to define strict anatomical boundaries and surgical logic. The front-end was designed to visualize these coordinates in real-time, bridging historical medical texts with a modern digital interface.

Challenges I ran into

Working solo in such a tight timeframe was a challenge. Mapping precise coordinates on the 3D curves of the human body using 2D images resulted in minor alignment deviations. However, I viewed this as a learning opportunity to understand the current limits of spatial AI in medical contexts.

Currently, the prototype achieves approximately 90% accuracy in point placement. However, mapping coordinates on a 2D image of a 3D human torso presents natural challenges due to anatomical curves. To eliminate minor deviations, the next phase of development will integrate Depth Estimation models to perform 3D mesh alignment, ensuring millimeter-perfect precision regardless of body posture.

Accomplishments that I`m proud of

I am proud of transforming a spontaneous spark of curiosity into a working prototype in just 96 hours. Seeing Gemini 3 successfully identify the 'Al-Kahil' point and align it with the spine was a major milestone for this project.

  • Surgical Precision: Successfully forcing the AI to recognize the human spine as a $X=50$ baseline for point alignment.
  • Holistic Dashboard: Creating a UI that feels like a premium medical suite rather than just a simple image generator.

What I learned

Building HacamaTech taught me that "Vibe Engineering" is just as important as hard coding. I learned how to guide Large Language Models to respect physical boundaries through descriptive spatial logic. Most importantly, I learned that a project’s strength lies in its ability to bridge the gap between human tradition and artificial intelligence. I learned that AI is not just about generating text; it’s about 'seeing' and 'reasoning.' This journey taught me how to guide a multimodal model to understand human anatomy and the importance of ethical, human-in-the-loop AI design.

What's next for HacamaTech

The journey is just beginning.

  • AR Integration: Bringing these points into an Augmented Reality (AR) environment for real-time practitioner guidance.

  • Real-time Vitals: Integrating wearable data to adjust Hijama points based on the user's blood pressure and heart rate variability.

  • Global Practitioner Network: Connecting HacamaTech users with certified professionals worldwide. To reach a clinical grade of reliability, I plan to fine-tune Gemini 3 with a dedicated dataset of over 10,000 diverse anatomical scans. This specialized training will enhance the model's ability to recognize subtle landmarks even in low-light clinical environments, making HacamaTech a robust tool for global practitioners.

  • Human-in-the-Loop for Safety & Ethics Patient safety is our absolute priority. Therefore, HacamaTech is designed not as an autonomous operator, but as a Clinical Co-pilot. The system suggests the most effective treatment points based on anatomical data, but the practitioner retains full control with a 'drag-and-drop' interface to adjust and confirm each point before proceeding. This ensures a 'human-in-the-loop' protocol for maximum safety.

Built With

  • agentic-workflow-design
  • gemini-3-flash
  • google-ai-studio
  • high-telemetry-surgical-ui
  • marathon-agent-protocol
  • o-lunar-intelligence-api(logic)
  • spatial-reasoning-engine
  • vibe-engineering
Share this project:

Updates