AI-Driven Project Scheduling Automation Tool

Advert

Freelance AI / Software Engineer required — UK Infrastructure SaaS

There is a planning process that sits behind a large slice of UK national infrastructure delivery. It is done by hand. It has always been done by hand. An experienced engineer spends two to three days on a drawing, a spreadsheet and a street map, and produces a programme of work that hundreds of people then execute.

Nobody has automated it. Not because the software is impossible — because the logic has never been written down. It lives in the heads of a small number of very experienced people.

I am one of them. I have spent the last year writing it down.

Where the project is now

A substantial amount has already been built for the MVP. The platform exists and works. What I do not have is an independent technical view on whether it has been built well — whether the architecture will hold, whether the decision logic is sound and testable, and what stands between where it is now and genuinely world-class software that can be taken to UK utility infrastructure giants and eventually worldwide.

What I need doing

A thorough review of the existing build, and a clear-eyed answer to one question: what would it take to make this world class?

The system covers:

Document understanding — extracting structured technical data from engineering drawings, mixed vector and raster, no two issuers formatting alike. The hardest part, and the part I most want a second opinion on.
A decision logic layer — expert judgement encoded into the system. My view is that most of this should be deterministic and testable rather than LLM-inferred. I want someone who will argue that through with me properly rather than reach for the nearest RAG pipeline.
Constraint-based scheduling — sequencing dozens of interdependent work areas against resource, geographic and statutory timing constraints.
Live data integration — several authoritative public and commercial APIs feeding the model.
Generated outputs — programmes, technical drawings and reports that go straight to a client. Presentation quality is part of the product, not a nice-to-have.
Multi-tenant SaaS wrapped around all of it.

I want to know what is solid, what is fragile, what is missing, and what you would do differently. Honesty is worth more to me here than reassurance. I am an engineer specialising in gas pipelines, and have no knowledge on AI / API / software development which is why I've had a company build it for me.

The deliverable is a written assessment covering the architecture, the extraction component, the decision logic and the test coverage, plus your view on the work needed to take it to world class. Fixed fee, agreed up front. If there is more to look at than that scope allows, tell me and we will agree it separately.

Why this is worth your time

The specification already exists. The domain expertise is on tap — you will be talking to me, daily, not to a product owner relaying answers. The commercial buyers are large, regulated, well-funded and currently doing this on spreadsheets. There is nothing like this.

If it is built correctly, it becomes the standard tool for the sector. The same model extends across several adjacent utility industries with near-identical constraints.

How I want to structure this

I should be straight with you about where I sit. I am the sole owner of the company. I work full time in the industry this product is built for — which is exactly why I know the problem is real and why the logic behind it exists at all. There is no board, no investors and no committee. Whoever comes in deals directly with me.

I want to do this in two stages.

Stage one — the review. A fixed fee, paid, for a bounded piece of work: look at what exists and tell me honestly what state it is in and what it would take to make it world class. I would rather pay properly for a few days of a specialist's time than ask anyone to work on spec.

Stage two — the build. If the review goes well and we both want to carry on, what follows is a partnership rather than a contract. Real equity, properly documented, in a new company set out to change how the utility infrastructure industry approaches design, planning and project delivery.

So: paid for the first piece, and a share of what it becomes if you want to build it. If you have wanted ownership of something with a clear market, partnering with someone who has key contacts in that market and works in it, and you have the technical depth to make it world class, this is worth a conversation. I'd like someone who wants to become a partner in the business, who is excited with the sound of the project and genuinely believes they can deliver a world class product.

How it would work

UK-based and English speaking. Not negotiable. I need direct, unambiguous conversation with the person actually doing the work, something I misjudged on the initial build.
Your own hours. Fully remote, work it around whatever else you have on. I care about the quality of the assessment, not when you do it.
Evening Teams calls. I am on site during the day, so evenings suit me for walkthroughs and questions. I will make myself properly available — the domain knowledge you will need is all in my head and I am happy to give you as much of my time as it takes.
Curiosity about the domain. It is the whole moat, and it is genuinely more interesting than it sounds.

One condition before we talk detail

I have deliberately kept this advert vague. The value of what I have built sits in logic that has never existed as software before, so a signed NDA is required before the project is explained, before any access to the codebase, and before any discussion of the domain. That is not a formality — it is the reason there is no incumbent in this market.

If the shape of this interests you, please get in touch. I will send the NDA over, and once it is signed I will tell you exactly what it is.