The AI ecosystem has a gaslighting problem. The word “open” is used so pervasively that its meaning becomes confusing. Companies with proprietary walled-garden offerings call themselves open. Other companies release technology and assign “open” licenses with heavy usage limitations. Many of the AI models shared are only partially open, with no visibility into upstream data or processing.

This might be good enough for AI-end users looking for AI-generated results from general chatbot services. However, many companies looking to leverage AI in their operations and services for their customers want more control over their AI systems. They want to leverage their unique data as a differentiation, not give away their data to mainstream services. They seek to innovate with AI, not simply provide the same lowest common denominator available to everyone.

Ainekkok seeks to make the word “open” mean “open” again in the AI ecosystem.

“Open” should mean that developers can quickly adapt models to their needs.  They shouldn’t have to reverse engineer models.  They shouldn’t need to make guesses to be able to fine-tune, quantize, and run models on engines such as llama.cpp.

“Open” should mean that people can be certain about the data used to train AI Models. They shouldn’t need to prompt AI models to guess whether personal data or copyrighted content was utilized in training.

“Open” should mean that AI developers know exactly what’s in their AI stack. They shouldn’t have to test against black boxes and unreliable frameworks.

The freedom and flexibility that open source software provides has not changed. Ainekko seeks to restore open source as an industry standard for the AI ecosystem so developers can study their technology stacks, adapt them, and redistribute them freely.

We believe in and foster open source communities and practices.  It's not sufficient to compare your technology to Linux.  Companies need to release fully open source models and technologies and build developer communities around them.  Only then will we see the Linux effect. 

Ainekko is helping create and support an open source AI technology stack. Our goal is to drive more innovation in AI by making models and enabling technology to be freely usable. Our first efforts include building an open-source composable stack from inference to hardware, and to foster an interoperable ecosystem around it called AIFoundry.org.

AIFoundry.org is a community initiative initially sponsored by Nekko.ai to create an open source interoperable ecosystem for AI.  It’s a home for opensource projects related to AI and engineers who want to build them collaboratively. 

We foster discussions and collaborations between machine learning engineers and software developers.  We encourage a collaborative, transparent development process for our member projects, and clear licensing.

We welcome other companies to participate in AIFoundry.org activities and projects, introduce their own open source projects to the AIFoundry.org ecosystem, or contribute projects. Other companies are welcome to become sponsors of AIFoundry.org.

We see an opportunity to standardize the AI hardware and software stack through Transformers. Transformers are universal enough to cover a lot of use cases and simple enough to optimize hardware to optimize and accelerate them.

Yes, it is. Nekko.ai is developing an end-to-end enterprise-ready stack to run various Transformer-based models that are customizable for any use case.
‍
Please submit a Contact Us form to learn more or inquire about investment opportunities.