Why AirfoilLearner
We didn't make CFD faster.
We made it agile.
Every other player in this space is building a faster solver. We built a different workflow — one that runs before the solver opens, costs nothing to start, and reaches the engineers that enterprise tools have never served.
The problem
Engineers spend half their project waiting, not thinking.
One simulation run costs £150 in idle salary before a single compute invoice is counted. A 50-run optimisation project burns £7,500. A team of three doing the same project: £22,500 — gone before the best design is even found.
The result is conservative engineering. When feedback takes 48 hours, engineers batch their guesses, run once, and stop exploring. Slowness doesn't just waste time — it changes how people think.
£150
idle salary cost per CFD run, before a single compute invoice is counted
£7,500
to run a 50-run optimisation project — per engineer
£22,500
the same project costs a team of three, before the best design is found
What we built
The agile layer in front of CFD.
A surrogate model that runs in the browser in 1.9ms — no setup, no meshing, no licences, no waiting. Engineers test hundreds of designs in the time it used to take to set up one run. And unlike any other surrogate on the market, ours comes with a validity envelope: it tells you when to trust the result.
This is what agile software development did for product teams in the 2000s. We're doing it for engineering teams now.
1.9ms
Browser inference
No install, no licence, no IT approval. An engineer opens a tab and runs.
100s
Designs per session
Evaluate geometries in the time it used to take to set up one solver run.
±σ
Validity envelope
We tell you when to trust the result. No competitor does this.
£0
Procurement
An engineer can be running in minutes. No budget owner, no purchase order.
Our vertical
A market enterprise CFD tools were never built for.
Approximately 6,000 SME aerospace and renewables teams across the UK and EU. Teams of 2–20 engineers. No dedicated IT. No procurement budget. No six-month software evaluation cycle.
These teams have the same physics problems as the largest aerospace manufacturers — they just can't access their tools. Our beachhead is offshore wind and UAV design — sectors where iteration speed directly translates to competitive advantage, and where our first industry pilot has already validated the workflow.
- Offshore wind turbine design
- UAV and drone aerodynamics
- Aerospace SME component design
- Renewables R&D teams
- CFD consultancies serving any of the above
£200M
Serviceable obtainable market
£2–5B
Serviceable addressable market
£25B
Total addressable market
Competitive difference
They are making CFD faster. We are making it agile.
A faster solver still runs once and waits. We run thousands of times before you open your solver. The comparison is wrong — that is the point.
Enterprise incumbents
Large simulation software vendors sell enterprise licences with IT procurement, purchase orders, and security reviews baked in. Their sales cycles run 6–18 months. Their smallest customer has more engineers than most of our customers have employees. The SME engineer never gets through their sales funnel.
We land in a day.
ML surrogate platforms targeting OEMs
These tools require a dedicated data engineering team on the customer side, minimum six-figure contracts, and custom model training pipelines. They are built for organisations with hundreds of engineers and a dedicated MLOps function.
We serve every customer they will never reach.
Developer-first frameworks
Some players offer toolkits for ML developers who want to build their own surrogate pipeline from scratch. Their user writes Python. Our user is a mechanical engineer who wants to run simulations, not write code.
Different buyer, different job, different product.
Well-funded deep-tech startups
Several well-funded competitors target large R&D teams only, with developer-first APIs and no browser product. No SME motion. No zero-procurement path.
We own the market they don't want.
Our four structural advantages
Distribution they can't replicate
Browser-first, zero procurement. An engineer runs in minutes with no IT sign-off. Incumbents can't undo six-figure licence structures.
The validity envelope
We don't just give a number — we tell you when to trust it. No SME competitor has this. It is the difference between a toy and a tool.
Pilot-driven accuracy credibility
Our first industry pilot produces validated accuracy data on a real-world application against real measured data. That becomes the sales weapon for everything that follows.
B2C engineer motion
We reach individual CFD engineers directly through domain networks. By the time a competitor looks down-market, we own the relationship.
Go-to-market
Three phases. One direction.
Land and prove
Close the current industry pilot as a paid engagement. Use real-world data to produce a one-page accuracy benchmark — specific error numbers, real engineer quotes, validated against measured data. Convert the existing waitlist. Get the first 5 paying subscribers at £80/mo.
Expand through domain networks
University and alumni networks for warm introductions into aerospace SMEs. Apply to clean energy and aerospace accelerator programmes. Target CFD consultancies as distribution — one consultancy using AirfoilLearner on every client project is worth more than 50 individual subscriptions. Activate international connections for a second named pilot outside the UK.
Raise
One paid pilot with published accuracy data. 20+ paying subscribers. Two or three domain-credible advisors. Warm introductions into the deep-tech investor community. Target £300k–£500k pre-seed. Use of funds: one senior CFD engineer or advisor, expansion into the next geometry domain beyond airfoils.
Traction
The numbers so far.
80+
aerospace engineers interviewed
60+
on the waitlist, growing 200% month on month
50,000+
simulations in training corpus
1
named industry pilot confirmed
Run your first section before you read the docs.
No meshing. No licence. No setup. Open and free.