Hi everyone, we’re Tyler, Diego and Trey — the team behind Tennr!

TL;DR We’re making it easy for non-technical businesses to bring custom, fine-tuned LLMs into their business.

🧐 The problem

Today, teams choose between using simple, ChatGPT-based AI use-cases or investing heavily in their own ML teams to build their own models.

If you’re building an LLM from scratch, you have to:

1. Find clean, expansive data
2. Generate thousands of prompt-response pairs
3. Use RLHF to improve outputs
4. Host your own GPUs
5. Build your own analytics
   
   Engineers and IT teams need to stay focused on their core functionality and can’t risk 6 month long dev projects that might not even result in a well-performing LLM.

💡 Our solution

Tennr is a platform that makes it easy to train, tune, and deploy custom Large Language models without needing support from engineering.

💥 How it works
We use proprietary techniques to turn a company’s document set into clean data that can train an LLM. Then, leveraging open source foundational models, Tennr automatically fine-tunes to support your ideal use-case. Finally, through Tennr, non-technical teammates can improve responses, make corrections, and monitor how their LLM is being used.

📚 Our backstory

Trey, Diego, and Tyler all met at Stanford.

• Diego and Trey competed together for 4 years on the varsity crew team.
• Tyler and Trey met as freshman roommates where they started an iOS consulting firm to get better at building apps
• All 3 studied Computer Science and focused on AI, design, and systems.

❤️ Our ask

• If you lead a business looking for a more custom AI solution and want to see Tennr in action, grab time on our site!
• If you like what we are building, share and upvote!