1 Week to Build the Future of AI with Humiris

Dear technology enthusiasts,
Since 2023, we’ve been working toward a dream. Not a distant, unattainable vision, but a tangible one: to provide AI engineers, developers, and builders with the tools to build AI that reflects their values and needs.This dream is Mixture of AI (MoAI).
MoAI is designed to simplify the lives of engineers, optimize AI infrastructure, and push the boundaries of what’s possible with today’s models. And to make it happen, we need you.
To celebrate this milestone, we’ve prepared a series of releases and resources for you!
‍
🌍 Building the Dream Together
‍
Our vision is bold: to create an intelligent, sustainable, and sovereign AI infrastructure. One that allows everyone to build their own AI model tailored with their data and values on a secure platform.
With your feedback, ideas, and support, we aim to bring together a community of committed builders, not only here in Silicon Valley but also across Europe and beyond.
To mark the launch, we’re hosting a week full of events, discussions, and demonstrations to collaborate with you and refine our vision together.
‍

🚀 LAUNCH WEEK PROGRAM
‍
📆 Monday: Mixture of AI – Basic Mode

Features: Explore Humiris Routing, our intelligent routing system that dynamically selects the best LLM based on your context and preferences. With Humiris Routing, reduce costs by up to 99%, improve quality, and minimize errors. Choose the LLMs you want to mix and just copy-paste the API key into your code, replacing OpenAI's with a more efficient mix-model.
You can find some resources here on how to get started with Humiris Routing.

📆 Tuesday: Monitoring
‍
Features: Presentation of the centralized dashboard that lets you configure and manage all your models from a single interface. Monitor, adjust, and analyze in real-time. This feature is for cost, speed, energy spending monitoring.
A Hacker News article introducing Humiris Routing. Join the conversation to share your ideas and feedback!

📆 Wednesday: Mixture of AI – Advanced Mode

Features: Discover Humiris Reasoning, a solution that builds personalized reasoning models using multiple LLMs, outperforming high-end single models like OpenAI-o1. Solve complex problems, adapt architectures, and achieve accuracy unmatched by any other model.
Official launch of Mixture of AI on Product Hunt. Test, vote, and share your thoughts!

📆 Thursday: Fine-tuning & Function calling MoAI

Features: Dive into advanced tools like fine-tuning, specialized assistants, and function calling to customize models to your specific needs. You will find a comprehensive tutorial on how to adapt Mixture of AI to your specific use cases, enabling you to fine-tune the Gating Model on how to select LLMS and better organize the reasoning patterns.
Open discussion on Hacker News about Humiris Reasoning.
Join us to explore the potential together.

📆 Friday: Introducing HumiChat

Features: Launch of HumiChat, a simple interface for interacting with Humiris Reasoning. A chance for everyone to test our advanced model!

📆 Saturday: Humiris Launch PartyEvent: AI Builder & Infra: Humiris Launch Party
Join us in San Francisco for a night celebrating AI builders and those committed to creating ethical and responsible AI infrastructure. Expect demonstrations, discussions and delicious food !

Join us here if you are in SF!

Humiris routing is available

For any request, Humiris Routing intelligently selects the most suitable model, ensuring optimal performance that surpasses standard foundation models across key benchmarks.

Try it now

See you tomorrow for Day 2 news! We’ve prepared lots of surprises for you.