Private preview

Not open yet.

Eclipse is in private preview. Enter the word a friend gave you to take a look around.

Eclipse · By invitation

Now booking Q3 2026 engagements

Built by operators.
Installed in your company.

Eclipseturns mid-market companies into AI-native ones — designed and run by engineers who've shipped at NASA, Vercel, Shopify, WordPress, and Nielsen. Not a deck. Not a framework. An installation.

— SHIPPED AT —
NASA▲ VercelShopifyWordPressNielsenMicrosoftLinkedIn
Why now

A 10% productivity lift on a 50‑person eng org is worth ~$5M/year.

Most companies are leaving that on the table because their teams are using AI like a search engine instead of a force multiplier. The gap between "we have Claude licenses" and "our team ships meaningfully faster with AI" is where Eclipse lives.

— OUTCOME · 01
3—5×
Faster cycle time on routine-ish work once Claude Code is properly installed in a workflow.
MEASURED · POST-ENGAGEMENT
— OUTCOME · 02
Day 1
Every engineer ships a real PR with Claude on the first day of the engagement.
RECEIPT · COHORT 03
— OUTCOME · 03
30 days
From kickoff to a forward roadmap you can hand to leadership without a translator.
PROTOCOL · TIER 03
The engagement

Three ways to install it. All built from the same protocol.

Every engagement runs the same playbook — audit, deploy, practice, scale. The tier sets the depth, the team size, and how much we customize. Not sure which fits? That's what the call is for.

01— 1 WEEK

Eclipse Spark.

A focused pilot for small teams ready to move fast. Onsite kickoff, full tooling setup, custom agents tuned to your codebase, and a 30-day forward plan. The fastest way to see what's possible.

BUILT FOR
5—25 person teams · single-department pilots
03— 30 DAYS

Eclipse Protocol.

The full installation. Org-wide fluency, custom agents built across departments, internal champions trained to run the protocol after we leave, business outcomes measured and reported to leadership. For companies committed to becoming AI-native.

BUILT FOR
75+ person companies · full transformation
Inside the protocol

Four phases. Compressed or
expanded to fit the tier.

— PHASE 01 / AUDIT

Audit

We embed with your team, read your codebase, and shadow your workflows. You get a current-state map, capability gap analysis per role, and a ranked-opportunity backlog.

SPARK
1 day
SPRINT
3 days
PROTOCOL
1 week
— PHASE 02 / DEPLOY

Deploy

The right stack, configured for your stack. Claude Code, Cursor, Cedar, MCP integrations, custom skills built against your conventions, SSO and security handled. By end of this phase, every engineer is shipping production work with AI in the loop.

SPARK
2 days
SPRINT
4 days
PROTOCOL
1 week
— PHASE 03 / PRACTICE

Practice

Role-specific intensives, hands-on builds inside your real codebase, daily office hours, and an internal #ai-wins channel we seed and run. This is where excitement compounds and adoption locks in.

SPARK
2 days
SPRINT
1 week
PROTOCOL
1 week
— PHASE 04 / SCALE

Scale

Custom agents embedded into the workflows that matter, measurement harness wired to real business outcomes, internal champions trained to extend the protocol after we leave. You get a forward roadmap and an exec readout.

SPARK
roadmap only
SPRINT
lite
PROTOCOL
full
What you keep

A library of Claude Code agents we built, battle-tested, and ship to every engagement.

One-command install into any repo. Pick a specialist, drop it in, get to work. Your engineers get the full library on Day 1 of any tier — and we build custom agents specific to your codebase during the engagement.

Browse all 6 agents
Why us

We're the team that ships, teaching the team that ships.

— 01 / WHO

Operators, not consultants.

The people running your engagement have shipped at scale. We've architected commerce platforms, built AI products in production, and launched apps used by millions. We teach AI fluency because we use it every day to build real things.

— 02 / HOW

Excitement over mandate.

AI isn't like a migration — we engineer the moments that make engineers want this. Day 1 wins, custom agents that solve your team's specific complaints, a tiny win-channel that runs itself. Surprising adoption is wins. Top-down adoption is the average of three.

Next step

Thirty minutes.
No pitch deck.

We'll ask about your org, your current AI usage, and what you'd want shipped in your engagement. We'll recommend a tier, scope the work, and send a proposal within a week. If we're not the right fit, we'll tell you who is.

CALIBRATED PER ENGAGEMENT · REMOTE + ONSITE · Q3 2026 · UNITED STATES