Our project minimum is $25,000.
Yes! Through prompt engineering, fine tuning, and retrieval augmented generation (RAG) we can create tailored solutions which generate insights from your data, not just the data an LLM was pre-trained on.
We're model agnostic, so everything from GPT-4 or Claude to LLama or Mixtral. We'll help you select the best model for your needs, including performance considerations and total cost of ownership.
In addition to preliminary consulting based on our experience and industry knowledge, we take an agile or "lean" approach: explore, refine, ship, and iterate quickly. Tight development cycles allow us to pivot as necessary in this constantly changing ecosystem.
We recommend development paths that take advantage of the excellent open source work currently available at all layers of the stack. Where necessary, we build custom solutions -- but only where necessary.