Skip to content

About Nautili

Nautili gives teams one place to route, monitor, evaluate and govern the AI they use.

Most teams no longer run AI through one model and one API. They use several providers, agents and internal AI tools, spread across different teams and people — and the controls around them end up just as scattered.

Why Nautili exists

As AI moves from experiments into production, the infrastructure around it fragments. Teams end up with different providers, agents, keys, policies and monitoring tools — but no single place to see or control what is happening.

Routing, cost, access, observability, evaluation and governance each live somewhere different, or nowhere at all.

Nautili exists to bring that control back into one layer.

Built from the problem

Nautili started as an internal solution.

We needed visibility and control over our own AI usage — where the spend was going, which models were being used and how requests were being routed.

Then our customers began running into the same fragmentation. That made it clear the problem was broader than one internal use case: teams needed a control layer for production AI.

  • Spend spreading across providers
  • No single routing layer
  • Limited visibility into usage
  • Governance getting harder as adoption grew

What we’re building

One control layer for every model, agent and AI workflow.

Nautili is growing beyond model routing into the control layer around production AI. Today that means routing, observability and governance in one place.

As AI systems become more agentic, that layer should also evaluate how models and agents perform, and govern the agents and people using them. That is the direction we are building towards.

Today
  • Routing

    Send every request to the right provider and model.

  • Observability

    See usage, cost and latency across teams in one place.

  • Governance

    Control keys, access and policies from one layer.

Direction
  • Evaluation

    Understand how models and agents perform in production.

  • Agents, and the people using them

    Extend the same controls to production agents and to a workspace where people work with AI.

Built by DevZero

Nautili is built by the DevZero team, with the same infrastructure-first approach to developer tooling.

Visit DevZero (opens in a new tab)

The team behind Nautili

We’re a small team building the control layer we wanted for our own AI infrastructure.

  • Debo Ray

    Debo Ray

    Co-Founder & CEO

    Founder; pricing, legal, go-to-market

  • Zvonimir Tomesic

    Zvonimir Tomesic

    Technical Staff

    Leads Nautili product and engineering; gateway, routing, partnerships

  • Anirudh Gupta

    Anirudh Gupta

    Product Manager

    Product, website content, onboarding, guides

  • Kron Raufi

    Kron Raufi

    Product Designer

    Website and product design system

  • Renato Nouman

    Renato Nouman

    Software Engineer

    Marketing site, app design system, analytics

  • Fran Leustek

    Fran Leustek

    Software Engineer, Harness

    Harness app and TUI

  • Ante Petrovic

    Ante Petrovic

    Software Engineer, Gateway

    Providers, model catalog, Bedrock, CI and publishing

  • Parth Agarwal

    Parth Agarwal

    Software Engineer

    Releases, billing, chatbot

Bring your AI traffic under control

Talk to us about your models, agents, providers and how your team is using AI.