Products ยท early access
Two products, shaped with early users
Elsor Router and Elsor Harness are not generally available. If one matches a problem you have, request early access and we will tell you what is ready.
Elsor Router
An intelligent LLM router that picks the best model for each request, balancing cost, latency, and quality.
The problem
Sending every request to one model means paying for power you do not always need, or settling for answers that are not good enough on harder tasks.
Key features
- Chooses a model per request, not per application.
- Lets you set how much weight cost, latency, and quality each get.
- Sits between your application and the models you use.
How it works
- Your application sends a request to the router.
- The router weighs the request against your cost, latency, and quality settings.
- It sends the request to the model it judges best, and returns the answer.
Elsor Harness
An intelligent AI agent harness, customized for a user or a team, so the agent works the way they work.
The problem
A general-purpose agent does not know your tools, your rules, or how your team likes work to be reviewed. People end up repeating the same setup and corrections.
Key features
- Configured per person or per team.
- Defined tools and permissions the agent may use.
- Team conventions and policies built in.
- Clear points where a person reviews the work.
How it works
- We learn how the person or team works and what tools they use.
- We set up the harness with those tools, policies, and review points.
- The agent runs inside that setup, and the setup improves as you use it.
Request early access
Tell us which product you are interested in and what you would use it for.