Solutions
One team for the roadmap, the build, and the proof it worked.
We can be your end-to-end partner for the AI roadmap and its implementation, or run the roadmap and hand it to your team, or another firm, to build. Either way, you get a prioritized plan before you get a bill for building anything.
Two ways to work with us
End to end
We run all three phases — the roadmap, the build, and the hand-off. One accountable team, from diagnosis to production.
Roadmap, then you build
We run the consulting phase and hand you a prioritized, costed roadmap. Your team, or another firm, builds it. You keep full optionality on who implements.
Custom pricing, sized to the engagement
Price depends on company size and complexity. End-to-end engagements have run from $10,000 to $300,000 — faster and better than a Big Five consultancy.
Every engagement starts with a free consultation call.
How we work
Three phases. Take one, or take all three.
01Consulting
Find out where AI actually pays off, and who should build it.
- The problem
- Most companies can't tell whether they're using the right AI tools, whether their team is using the ones they have well, which workflows are actually worth automating, which agents are worth building for employees, vendors or customers, which SaaS could be replaced by something cheaper and better.
- How we find out
- We interview leadership, then put an AI voice agent in front of everyone in the company. It learns how people actually work today — their workflows broken into steps, the tools and software they use, and how they use them.
- What you get
- A prioritized AI roadmap across AI tools, automation and SaaS replacement, ranked on impact against feasibility, each with a timeline and a price. It names the builders inside each team, so we can train them before we train everyone else, and ties every recommendation to the value it's expected to create.
02AI Enablement and AI Development
Build it, and get the team using it.
- Training
- Hands-on training on the tools and agents you already have and the ones we build, plus general AI literacy for the wider team. We also handle the internal and external communication that makes the change land, instead of stalling on a team that's resistant to it.
- Engineering
- Automation built inside your existing AI tools, or as standalone software when vendors or customers are in the loop. We either connect your SaaS to a new system of record/understanding, or we replace your SaaS with something built for exactly what you need.
03Hand-off
Prove the return, and leave your team able to run it.
- The number
- We close the loop on the value hypothesis from the consulting phase and finalize the actual return on investment.
- The team
- We train the people responsible for keeping the new systems running, so nothing depends on us staying in the room. We stay for as long as you need us.
Why Eclipse
Proximity beats magnitude.
A total solar eclipse happens because the moon is about 400 times smaller than the sun, and about 400 times closer. Proximity beats magnitude, exactly — and that is the whole argument for Eclipse against a global consultancy. We are far smaller, and we sit far closer to the work, and the effect is identical. The incumbents are the sun: enormous, distant, and impossible to engage with directly. Being small is not the caveat in our story. It is the mechanism.
Operators, not slide decks.
Our team is ex-Google, NASA, LinkedIn, Vercel, Casper Studios (acquired by Ode,with Anthropic), Shopify and Nielsen. We're technical solution engineers — operators and builders who've shipped AI inside Fortune 100 companies and inside pre-seed startups alike. That range is the point: we've done this at both ends of company size, not just in the deck.
Advisors from the labs building the models.
Our board includes advisors who lead teams at Anthropic and OpenAI. That seat gives us an early, working read on where model capability inside the enterprise is actually headed, not where a roadmap says it's headed. It's also why we optimize for ROI first: we minimize your spend on tokens needed for the maximum impact, rather than burning compute to look sophisticated.