Overview / Usage

We're producing the **enterprise backoffice** of the future using SCM on self-hosted git repack(s) where teams of ai agent teams produce enterprise analyses and code automatically & with no new interface for the user.

​

Ultimately someone with no coding knowledge should be to deploy the most powerful automated backoffice or component of an office, whether accounting or coding or legal or marketting or business operations.

​

We're so confident this is the future of business operations we're developping a sustainable open-source business model for a revenue business.

Methodology / Approach

We're producing a generalist / specialist agent for enterprise & business operations.

• to that end we're using a "compile" application approach where the application re-compiles itself based on user needs every new deployment.

• compile composes pipelines , prompts, and deployments of standalone webservers for code based on specialized evaluation and diverse llm/ml models , each of which is selected and fine tuned.

• pipelines are agent team of teams on graphs that communicate securely over self managed git and (multi)cloud.

• certain agents can deploy their own stack , then use it async in an optimized way.

• we're committed to accessibility in more ways that one, and that translates to the fact just about any user can deploy powerful and complex systems in their own language with a single free text "command"

Technologies Used

• Cool Intel Docker Images for ML
• Unit Test
• MultiModal

• optimum-intel
• huggingface.co/intel
• Open VINO

Other links

• Contribute : Source Code
• Huggingface