SHIPdock

Logo files

  • Full logoFor dark backgrounds SVG PNG
  • Square logoAvatar and app icon SVG PNG
  • Full logo, blackFor light backgrounds SVG PNG
Start with an auditStart audit

Your company’s safe passage to being AI‑native.

SHIPdock is an AI implementation partner. Our engineers work inside your organization to find where AI saves real time and money, build the fixes with your own people, and prove the results.

Start with an audit How it works

3 of 5 companies filledLimited intake

  • 4-week audit
  • Paid AI roles
  • Everything belongs to you. No license fees.

Most companies bought AI tools. Few know how to use them, or where they pay.

The tools are paid for, but people mostly use them to rewrite emails. Nobody knows which work AI should change. Consultants are slow and talk only to managers. And staff who fear for their jobs don't describe how they really work, so every plan starts from wrong answers.

  • Rep Reported, never verified

    “We use AI.” Mostly to rewrite emails.

  • ED Existence doubtful

    The return on AI tools nobody really uses.

  • PA Position approximate

    An AI plan made with managers only.

  • PD Position doubtful

    How the work is really done. People who fear for their jobs don't say.

  • Unsurveyed

    Which of your apps AI agents can even connect to.

40% of employees worried about losing their job to AI in 2026, up from 28% in 2024.

Mercer, Global Talent Trends 2026

Rep, PA, PD and ED are the abbreviations nautical charts use for hazards nobody has verified.

Audit. Build. Grow.

The audit is the door. Every round has to prove its savings before you buy the next one.

In difficult water a ship takes on a pilot: someone who knows the channel comes aboard and steers alongside the crew. Our forward deployed engineers work the same way, inside your company and next to your people.

Top view of a ship with a pilot launch alongside: the pilot boards and steers with the crew. 090° TRK-017 · YOUR COMPANY FDE BOARDS HERE
Step Leg What happens What you get How long

Audit

Map your work and your apps.

Your people time the jobs they repeat in shipdocs.ai, our audit app. At the same time we audit every app your company uses and pays for: how AI-ready it is, and whether AI agents can connect to it.

Where time and money go, an AI‑readiness check of every app, a roadmap of 3 to 7 fixes, and paid AI roles ready to fill.

 About 4 weeks

Build

Fix it with your people.

Our engineers build the fixes together with your AI Heroes. Then the same people time the same jobs again.

Fixes running on your own servers and in your own tools, with savings measured against the audit baseline.

 2 to 3 months

Grow

Keep the change inside.

New audits, proof reports, support for your AI Champions and AI Heroes, and AI training for staff.

The next round of fixes already mapped, and a team that runs what we built.

 Monthly

Grow feeds the next audit.

Everything we build belongs to you.

Every fix, agent and connection runs on your own servers and in your own accounts, not on ours. You pay for the work, not for a license, so there is no monthly fee to keep using what we built.

Runs on
Your servers and your accounts
Owned by
You, from the first fix
License fees
None

Four weeks to a roadmap you can prove.

Every audit runs the same five stages. One rule is hard-coded: nobody times a single job until the trust stage is done.

  1. 01

    Setup

    Admin, with our FDE

    Teams, people, costs and apps go in. Every app gets a connection check.

  2. 02

    Trust

    Organization owner

    Video, written promise, paid AI roles and rewards are prepared, reviewed and published.

  3. 03

    Tracking

    Employees

    Each person times 5 to 10 jobs they repeat and answers a 2‑minute daily check-in.

  4. 04

    Results and roadmap

    Owner, with SHIPdock

    Jobs ranked by cost, 3 to 7 fixes chosen, paid AI roles filled.

  5. 05

    Fix and prove

    AI Heroes, with our FDE

    Fixes built and launched, jobs timed again, pay released.

The launch lock

Tracking locked

The launch button stays locked until every item is done.

  • NDA and data processing agreement signed
  • Leadership video uploaded
  • Promise signed by the owner
  • Paid AI roles set, with real pay
  • Rewards set
  • Privacy notice approved
  • Employee rep review done

Everyone on board wins.

AI only pays off when every side comes out ahead. So every audit is set up as a deal: your people and SHIPdock get paid for their part, management saves money on the work that gets automated, and nobody has a reason to hide how the work is really done.

The deal, drawn as a plan view of your company as a vessel. Management is on the bridge, your people are on deck, and SHIPdock's pilot launch is alongside. White lines carry trust: the main trust line runs from the owner to your people, and SHIPdock's lines run to the owner and to your people. Green lines carry money: paid AI roles and rewards from the owner to your people, a fee for proven savings from the owner to SHIPdock, and savings from automated work back to management. TRK-017 · YOUR COMPANY 090° Management BRIDGE FIXED FIXED NEXT ON DECK Your people SHIPdock PILOT LAUNCH · OUR FDE VIDEO · SIGNED PROMISE METHOD · TOOLS · RULES NOBODY SEES ONE PERSON’S DATA PAID AI ROLES + REWARDS FEE FOR PROVEN SAVINGS SAVINGS FROM AUTOMATED WORK The deal, drawn as a plan view of your company as a vessel. Management is on the bridge, your people are on deck, and SHIPdock's pilot launch is alongside. White lines carry trust: the main trust line runs from the owner to your people, and SHIPdock's lines run to the owner and to your people. Green lines carry money: pay and rewards from the owner to your people, a fee for proven savings from the owner to SHIPdock, and savings from automated work back to management. 000° Management BRIDGE ON DECK Your people SHIPdock OUR FDE PAY FEE SAVINGS

Main trust line. The owner explains why, signs the promise and puts real pay on the table.

SHIPdock’s trust lines. We bring the method, the tools and the rules, and we guarantee that nobody at your company ever sees one person’s data.

Money. Pay and rewards for your people, our fee for proven savings, and what management saves by automating.

Management

The owner and the leadership team.

FearsNo clear AI plan, staff pushback, money spent on tools nobody uses, and falling behind.

Saves money by automating repeated work.

Example One fix in our example saves about 245 hours, or about 5,900 EUR, a year.

Also gets
  • A proven roadmap in about 4 weeks
  • Staff on board
  • Everything we build, on your own servers, with no license fees

Your people

Employees and team leads.

FearsBeing replaced, being watched, and nothing in it for them. Team leads fear being blamed for slow work.

Paid for taking part.

Paid AI roles: a monthly allowance for AI Champions, role pay and a results bonus for AI Heroes. Plus rewards.

Also gets
  • A written promise, signed by the owner
  • Their most annoying work, fixed
  • For team leads: team-level results only, never individual ones

SHIPdock

Us: the engineers we send in.

FearsBad data from scared people.

Paid to deliver measured savings, not slides.

You pay for the work, not for a license.

Also gets
  • Honest data
  • A happy client
  • The next project

Trust before data.

People who fear for their jobs don't describe their real work. So before anyone times a minute, the owner stands in front of the team, explains why, and puts promises and real pay on the table.

15% to 55% of employees feel positive about AI when leadership clearly supports it.

BCG, AI at Work 2025

What every employee sees first, in this order

  1. A video from the owner. Two to three minutes on why now, what it means for jobs, and what's in it for them.
  2. What we collect, and what we never collect. In plain words.
  3. The promise. Signed with the owner's name and date.
  4. Paid AI roles. How many, what they pay, how to apply, and when.
  5. Rewards. Real amounts and rules.
  6. A question box. Anonymous, answered in public.

Never collected

  • What people type
  • Webcam or microphone
  • Private emails or chats
  • Scores or leaderboards of people
  • One person's data, shown to their manager

Rules of engagement

Work, not people.
We measure jobs. We never score or rank a person.
Only while a job runs.
Nothing is recorded when no timer is running.
The employee is in control.
They pick their jobs and press Start and Stop themselves.
Groups of five or more.
No result is ever shown for fewer than five people.
Your data is yours.
Every employee can see, correct, download and delete their own data.
People decide about people.
Paid roles are chosen by people. Software never ranks applicants.
Short memory.
Run-level data is deleted 90 days after the audit ends.
Everything under NDA.
Signed before setup. EU privacy rules from day one, data stored in the EU.

You get new roles inside your company, with our help.

We help you create paid AI roles for your own people, and the people who help transform your company get paid for it. That's how the change stays inside after our engineers leave.

AI Champion

A paid part-time role inside a team.

Work
Helps colleagues use AI and new fixes, runs office hours, and is the team's voice to our engineers.
Time
2 to 4 protected hours a week.
Pay
A monthly role allowance on top of salary.
How many
At least one per team, roughly one per 25 to 75 people.

AI Hero

A paid role that owns one fix, from idea to proof.

Work
Builds a fix with our FDE, tests it with the team, and keeps it running after launch.
Time
Often 20 to 50% of the week while the fix is built.
Pay
Role pay while building, and a results bonus when the fix is proven.
Path
A permanent AI job offer after one or two proven fixes.

63% of employees would trade a 10% raise for chances to build AI skills.

Mercer, Global Talent Trends 2026

Can AI agents actually work in your apps?

A costly job inside a closed app is much harder to fix. So every app you use gets a connection check, not only a time study.

Illustrative app connection map: which apps AI agents can reach, and how many hours of work sit in each. Agent-ready: Gmail, 41 hours a week; Outlook, 23; Pipedrive, 22; Xero, 37; Google Sheets, 17. API-friendly: Excel, 24; delivery system, 18. Export only: warehouse app, 26; payroll, 6. Closed: legacy ERP, 33.
Illustrative map. Your audit produces the real one.
Agent-ready
The app has an MCP connection. AI agents can work in it right away.
API-friendly
It has an API but no MCP. Our engineer builds the connection first.
Export only
Data leaves only as files. Fixes are half-automatic, with a person moving the files.
Closed
No API and no export. Often the best fix is moving to a more open app.

MCP, the Model Context Protocol, is a standard plug that lets AI assistants work inside an app. An API is a door an app opens for other software.

One engagement, from the first video to a proven fix.

Follow one example company through an audit and its first fix, week by week: who goes first, what gets timed, who gets paid, and what it proves.

An example engagement. Names, times and numbers are made up.

  1. Week 0, setup

    The company goes in.

    Teams, people, costs and apps are set up with our FDE. Every app gets a connection check. Nobody times anything yet.

  2. Week 1, trust

    Mike, the owner, goes first.

    Mike records a 2 to 3 minute video: why now, what it means for jobs, and what’s in it for each person. He signs the promise and publishes the paid AI roles and rewards: 4 AI Champion roles and 3 AI Hero roles.

    Every employee sees the video first. Tracking unlocks only after all of this is done.

    • Lock7 of 7 done · tracking unlocked

  3. Week 2, tracking

    Sarah times one job.

    Sarah, an accountant in a team of 6, times “Enter supplier invoices.” She presses Start at 9:02 and Stop at 9:39. The 12 minutes she was away are snoozed and not counted.

    • 09:02Start · Enter supplier invoices
    • Away12 min · snoozed, not counted
    • 09:39Stop
    • shipdocsSaved: 25 minutes. 1 of 3 runs.

    In her day‑6 check-in she names the slow part, “Copying numbers from PDF invoices by hand,” and an idea: “Read the PDF and fill in the fields for me to check.”

  4. Weeks 2 to 3, tracking

    Every team times its own jobs.

    In Sales, Jake times “Write a price offer”: 41 minutes and 23 switches between apps. In Support, Lisa runs a count job, “Answer delivery questions”: 14 questions in 52 minutes.

    • SalesWrite a price offer · 41 min · 23 app switches
    • SupportAnswer delivery questions · 14 in 52 min

  5. Week 4, results and roadmap

    The roadmap names the fixes, and the people.

    “Enter supplier invoices” is one of the top jobs, and one paid AI Hero role opens to fix it. Sarah applies with her check-in idea. Mike chooses her: role pay while she builds, and a results bonus when the fix is proven.

    Jake becomes the paid AI Champion for Sales, with a monthly allowance and 3 protected hours a week. His colleague Emily becomes the AI Hero for the offer drafter.

    • AI HeroInvoice reader · Sarah
    • AI ChampionSales · Jake
    • AI HeroOffer drafter · Emily

  6. Weeks 5 to 12, fix and prove

    The fix goes live. The job is timed again.

    Sarah builds the invoice reader with our FDE. Once it is live, the 4 people who do the job time it again for 2 weeks. It drops from 25 to 9 minutes per run: about 245 hours and about 5,900 EUR a year.

    • Baseline25 min per run
    • Proof9 min per run
    • Formula(25 − 9) min × 5 runs × 4 people × 46 weeks ≈ 245 h × 24 EUR ≈ 5,900 EUR

    Support gets its fix too: Lisa’s 14 questions now take about 20 minutes.

  7. After the audit, proven

    Proven. Then the next round.

    The fix reaches Proven, and Mike approves Sarah’s bonus. After her second proven fix, she is offered a permanent role: AI Operations Lead.

    In his monthly update to all staff, Mike writes: “Support now answers delivery questions more than twice as fast.” Grow feeds the next audit.

    • StatusProven · bonus approved
    • OfferAI Operations Lead · permanent role

Proof, not decks.

We're paid to deliver measured savings, not slides. Your audit runs are the baseline. After a fix goes live, the same people time the same job again. The difference is the proof.

Enter supplier invoices

An example job. Move the numbers to see how proof is calculated.

25
9
5
4
24

Saved per run

16 min

Saved per year

245 hours

Worth per year

5,900 EUR

(25 − 9) min × 5 runs × 4 people × 46 weeks ≈ 245 h × 24 EUR

Illustrative numbers from our method, not a client result. Your baseline comes from your own audit.

The people who come aboard.

Two roles on our side. In a small engagement, one person can be both.

  • Edvarts Gaigals, Engagement lead at SHIPdock

    Edvarts Gaigals

    Engagement lead

    Owns the relationship with your company. Runs the kickoff and the monthly steering meeting.

  • Davids Petuhovs, Forward deployed engineer at SHIPdock

    Davids Petuhovs

    Forward deployed engineer

    Runs the audit day to day, picks fixes with you, builds them with your AI Heroes, and trains your AI Champions.

Start with an audit.

Tell us about your company and what you want AI to fix first, and let’s talk. Every price is individual, based on your industry, size and revenue.

  • About 4 weeks
  • Everything under NDA
  • EU privacy rules from day one
  • Everything we build belongs to you

3 of 5 companies filledLimited intake

What should the audit answer first?

Pick up to three.