OPEN TO WORK
Dry dock to open water.
Dry dock is where the work happens. Open water is production, not staging. We work alongside your engineers, write the plan before the code, and hand over software you can maintain without us — and we run our own products, Slate and Tack, on the same spine.
Studio manifest
- In production
- Slate v1.1.1
- In build
- Tack — dashboard + extension
- Last deploy
- Today
- Engagements
- 2 active
- Next availability
- Two build slots, Q4 2026
What we do
3 lines of work01 / Consulting
Consulting
Bring us in on a specific problem — architecture reviews, technical due diligence, unblocking a stuck team. We work alongside your engineers, not around them.
- Architecture review
- Technical due diligence
- Team unblocking
02 / Client projects
Client projects
Hand us a product to build end to end. We staff a dedicated team, ship on a clear timeline, and hand over code your team can maintain.
- Dedicated small team
- Weekly demo, no account manager
- Handover runbook included
03 / Our own products
Our own products
We design, build and run software of our own. It keeps us shipping to real users constantly rather than only billing hours on someone else’s roadmap.
- Slate — in production
- Tack — MVP, building
- One shared spine
How we work
SCOPE → BUILD → SHIPSCOPE
We start with a working session to define the problem, the constraints, and what "shipped" actually means. You get a written plan before any code.
Deliverable · Scope document, estimate, stated date
BUILD
A small team with direct access to the engineers doing the work. Weekly demos, no account managers relaying messages.
Deliverable · Working software, demoed weekly
SHIP
We deploy to production, not to a staging environment. You get working software and the documentation your team needs to run it.
Deliverable · Production deploy, runbook, handover session
Selected work
All work →Fintech / Payments ops
Ledgerline
40M transactions/day. 6hr close cut to 40 min.
2025Client project · 14 weeksPostgres · Go · Kafka
- Transactions reconciled per day
- 40M
- Month-end close, down from 6 hours
- 40 min
- Plan to production
- 10 weeks
Healthcare / Scheduling
Northfield Health
Zero-downtime cutover across 12 clinics.
2024Client project · 9 weeksPostgres · TypeScript · Temporal
- Clinics migrated
- 12
- Minutes of scheduled downtime
- 0
- Dual-write window
- 3 months
Labs we've shipped
Ours. Live. Maintained.Software we design, build and run ourselves — not client work, just things worth shipping.
Slate
Client and project management for freelance agencies — and the software Dockyard runs itself on.
- Tiers
- Platform vendor, agency, agency client
- Isolation
- Strict per-tenant, enforced in the query layer
- Hosting
- Self-hosted, long-lived Docker containers on a VPS
- Dogfood
- Dockyard runs its own agency on it
Ingot
Visual QA and website feedback: clients annotate live sites, the team triages the result.
- Shipped
- Dashboard, browser extension
- Roadmap
- Session recording, realtime collaboration, installable snippet
- Spine
- Shared with Slate — pnpm/Turborepo, Postgres, Docker
- Isolation
- Strict per-tenant
Shipped, not staged.
“Dockyard's team shipped our reconciliation rebuild in ten weeks. They wrote the plan, showed up to every demo, and left us with a codebase our own engineers could maintain.”
Ship with us
One form. One reply. One business day.Tell us what you’re building.
An engineer reads every one of these. If it isn't work for us we'll say so, and usually say who to ask instead.