Project HANSAL: Hybrid AI for Next-gen: Sustainable, Affordable, and Lightweight
Recent advances in AI demand significant computational resources, which presents cost and sustainability barriers for small and medium-sized enterprises (SMEs).
Project HANSAL directly addresses these challenges, offering a framework that balances accuracy, runtime, cost, and energy consumption.
This project is a framework that is also featured in the Master's Thesis @ FH Kiel: Study on Proof-of-Concept Benchmarking Framework for Resource-Optimised AI: Efficient Performance, Cost, and Sustainability for Small-Medium Enterprises (SMEs)
- Exploration and testing of AI benchmarking tools suitable for SMEs.
- Evaluation of currently available hardware architectures (GPUs, AI accelerators) for affordability and scalability.
- Implementation of a lightweight benchmarking workflow using open-source libraries (e.g.
CodeCarbon,Zeus,Perun). - Creation of an open-source repository for replicable benchmarking experiments, supporting business case-specific needs.
- Provide SMEs with the tools and know-how to adopt AI cost-efficiently and sustainably.
- Serve as both an academic contribution and an actionable industry guide.
- Structure data pipelines for transparent performance, allowing resource optimization tailored to SMEs.
- Literature Review & Data Collection: Oct 2025 – Dec 2025
- Benchmarking & PoC Development: Dec 2025 – Jan 2026
- Thesis Writing: Jan 2026 - Feb 2026
- Colloquium: Mar 2026
Copyright (c) 2025 Heansuh Lee. All rights reserved.
This GitHub repositoy, its associated software, the documentation and its content are proprietary to Heansuh Lee.
No part of this work may be reproduced, modified, or distributed in any form or by any means without prior written permission from the copyright holder.