Skip to content

LDN 51.5072° N

Recommendations
don't ship. Operators do.

Forward deployed engineering for maritime and B2B software. An operator in the room deciding what gets built, an engineering team that builds it, and something running in production when we leave.

PILOT TO FLEET · MARITIME & TRADE · EST. 2026 — LONDON

DIAGNOSIS / 01

Activity is rising.
Decision quality isn't.

420 organisations were active in maritime AI last year, up from 276. The budget is committed. The distance between buying it and trusting it is the work.

SYMPTOM / 01
Nobody owns the distance from pilot to fleet.
A pilot proves the product on one desk. A rollout needs workflow change across teams that didn't buy it. Implementation ends at go-live, and the superuser responsible for its success has a day job.
SYMPTOM / 02
One question, four systems.
Each tool does its job. The commercial question spans all of them, so the answer gets assembled by hand, under time pressure, by whoever is on shift.
SYMPTOM / 03
The information exists.
The interface doesn't.
Performance, compliance, commercial and operational data all sit somewhere. What's missing is one view at the moment the decision is made. Defining that view is a product decision. Building it is an engineering job. They have to be the same engagement.
20+ years operating, building and scaling businesses

WHO / 02

You're not buying a method. You're hiring the operator who has run it.

Lucid is led by Leigh Steed-Middleton, a product and technology executive who rewires operating spines rather than tweaking products. Across five years at Sedna he put product, engineering, marketing and customer success under one system: recurring revenue grew 10× at roughly 58% CAGR, six acquisitions were integrated, and the customer base passed 700 organisations.

Structure where there's noise, velocity where there's friction, accountability where there's ambiguity. The operator you meet stays on the work. Behind them, a small engineering team — so the seat ships, rather than writing a plan somebody else has to resource.

Read the background ↗

MODES / 03

Two modes. The second is what makes the first ship.

The same people who run the model, holding the keyboard.

— 01 .ENGAGEMENT .MODE 01 > TAKE THE SEAT .FORWARD DEPLOYED
® LUCID (...) OPERATING ...... »»»»»»
(INSIDE YOUR LEADERSHIP TEAM) (OPERATE)
// — PART-WEEK / EMBEDDED
= 01001

We take the seat.

Fractional CPTO / COO — and we take the seat inside the work. Customer calls, the roadmap, the architecture, the hiring. What we decide, our engineers build.

Practice ↗
— 02 .VENTURE .MODE 02 > MAKE THE THING
® LUCID (...) SHIPPING ...... »»»»»»
(PRODUCT, ARCHITECTURE, DELIVERY) (BUILD)
// — IN-HOUSE / SHIPPED
= 02002

We build the thing.

Some problems don't need advice. Lucid incubates and operates its own software, and takes build work where the product is the point rather than a deliverable. The same capacity sits behind the seat, which is why an engagement can end with something running — and why it costs what it does.

Evidence ↗
Mode What we own Shape Status
Fractional CPTO Product, engineering, architecture, technical hiring Ongoing, part-week OPEN
Fractional COO Operating cadence, delivery, spend, cross-functional flow Ongoing, part-week OPEN
Operating review Written assessment of the model, and the plan we then run Fixed scope OPEN
Build Software designed, built and run by our team Project or embedded OPEN
The full practice ↗

PROOF
“There is a noteworthy difference between simply connecting systems and confirming
they actually support decision-making.”
THETIUS & DANELEC · THE GREAT INTEGRATION · 2026

Confirming it is the job we take.


METHOD / 04

How an engagement runs.

Four phases. The last one is the point — we're building something that keeps running without us.

01
Assess
We read the machine, not the deck. Delivery flow, decision rights, spend, architecture, the people who actually unblock things. You get a written operating review.
02
Install
Objectives, decision rights, the delivery cadence and the small number of metrics that are allowed to matter. Written down, agreed, and in use within the first month.
03
Operate
We hold the seat. Hiring, vendor calls, architecture, board material, the hard conversations. Where the build is the constraint, our engineers ship alongside yours.
04
Hand back
A permanent leader, briefed. A model documented. A team that runs the cadence themselves. We stay reachable; we stop being necessary.

FIT / 05

Who we work with.

FIT / 01
Founder-led, scaling
Post-fit, pre-structure. You need a technology and operations leader before you can justify hiring a permanent one.
FIT / 02
Portfolio operators
PE and VC teams who need a real operator inside a company — a turnaround, or a gap between CPOs, CTOs and COOs.
FIT / 03
Enterprise leadership
A division that has to move like a company. We work inside the constraints you actually have, not the ones a consultancy assumes.
FIT / 04 PILOT TO FLEET
Software companies with a deployment problem
The product works. Customers can't get it live. The distance from pilot to fleet is a real job, and at most vendors it belongs to nobody. We own it, and we leave the pattern behind so the next rollout is cheaper. This is the forward deployed function most vendors are trying to hire.
THE RECORD / 06

What we install is written down. What we build is running.

An engagement leaves a record on purpose. The operating review, the decision log, the architecture position, the hiring plan — documents your board can read and your team can run from. Where we build, it leaves the software too: running in your estate, with the people who own it named. If it only lived in our heads, we didn't do the work.

OPERATING REVIEW DECISION LOG DELIVERY CADENCE ARCHITECTURE POSITION WORKING SOFTWARE

EVIDENCE / 07
ENGAGEMENTS — 04
String Ting
Technology and operating model, run by Lucid.
2026 —
LIVE
LADN.
Repositioned, repriced and rebranded with Lucid.
2026 —
LIVE
Sedna
Forward deployed: an operator in the seat deciding the software, a pod shipping it.
2026 —
LIVE
Aquila
Rebranding, repositioning and repricing with Lucid.
2026 —
BUILDING
VENTURES — 02
Reel
Nothing leaves a shared inbox unseen. Built and owned by Lucid.
2026 —
BUILDING
Undisclosed
A second product, in build.
2026 —
BUILDING

Six on the record: four engagements, and two products we build and own. Three live, three in build. We publish work at the point it's real, not at the point it's announced.


WRITING — THE PUBLICATION

Notes from inside the machine.

Operating models, product organisations, and the unglamorous mechanics of shipping software. Written by the operator doing the work, sent when there's something worth sending.

The one in progress is here for the same reason the ledger shows what is still building.

START

Forty-five minutes. Bring the constraint.

Tell us where the model is bending. We'll say what we'd do, and whether we're the right people to do it.

WHAT DO YOU NEED?
LONDON — GMT TAKING ENGAGEMENTS