Try it Yourself
LLM powered copilots
Easily build and deploy LLM-powered copilots that are experts on your platform using Layer's proprietary path abstraction.
How it Works
Deliver Users a Powerful Chat Experience
Users can now tell Layer Copilot exactly what they want to do through voice or text.
The LLM Goes to Work
User requests are fed into the LLM, which creates a path of functions that Layer Copilot executes to accomplish the user's goal.
Layer Copilot Improves Over Time
User feedback is collected on an opt-in basis and used to improve Layer Copilot's performance!
Try our live demo by visiting this website on desktop
Scroll to play animation
Try it Yourself
Our path abstraction
enables developers to build production-grade copilot applications in days instead of months.
The abstraction allows for type-safe execution of a chain of functions. This enables faster throughput, higher reliability, and a degree of simplicity that no other framework can offer.
Frequently Asked Questions
What is our company trying to do?
Layer was founded in 2023 with a mission to increase the productive output of every software interaction. We do this by providing the infrastructure to enable thousands of companies to integrate LLM-powered copilots into their platforms.
What is a copilot?
A copilot is a platform assistant users can interact with through voice/text. The technology underlying a copilot is an orchestration of specialized LLMs and vector databases that enable users to
interact with software platforms through natural language.
Isn’t a copilot basically just a chatbot?
The key difference between copilots and chatbots is that a copilot can act agentically and a chatbot cannot. From an end-user perspective, this means a copilot is able to take actions on their behalf, essentially functioning as a personal assistant with complete knowledge of the platform.
Why aren’t copilots already everywhere?
Building a copilot requires a company to give an LLM direct access to their product. This magnifies the potential consequences of unpredictable outputs from the LLM. Solving the reliability problem is impossible with existing general-purpose frameworks.
How does Layer solve for reliability?
The foundation of our framework is our path abstraction. The abstraction is backed by a proprietary sequence validation system that guarantees any user input that is typed into a copilot will be mapped to a sequence of functions that can be safely and predictably executed on a given platform.
How do I integrate a copilot with Layer?
Integrating a copilot with Layer is a two-step process. Our technology generates a structural representation of your platform by reading the DOM. This is used to provide a base degree of functionality out-of-the-box . We then provide a simple and standardized interface for deeper and more powerful integration.
How do I get started?
We plan to launch our self-serve model in Q4. Please fill out the form on our website and we will notify you when this becomes available. If you want to get started sooner, please contact firstname.lastname@example.org, and we will respond with next steps.
Layer is the first step toward standardizing reliable agentic communication with LLMs.
Join us for this journey.
Thank you! Your submission has been received!
Oops! Something went wrong while submitting the form.