The engineer you speak to builds the system.
Caius Seemann, Software and Machine Learning Specialist.
Experience.
I started building commercial systems at sixteen, running email campaigns for a B2B HR software company — a demanding audience to sell to, and a fast education in the difference between something that looks like it's working and something that actually is.
My background is mathematical — physics, and a lasting interest in particle physics. That sounds irrelevant to a plumbing firm's invoicing, and mostly it is, except for one habit it drills into you: you don't get to believe a result until you've tried to disprove it. A number that looks good is a starting point for an investigation, not a conclusion.
That habit is why this business works the way it does. It's why the first thing I sell is a measurement rather than a build, why some audits end with "don't automate this", and why nothing goes live until it has proved itself against your own team's decisions. I have built research systems whose entire purpose was to demolish my own best ideas — and they did. Being able to change your mind with evidence is not a soft skill in this work; it's the whole job.
Since then the work has narrowed deliberately: taking manual, repetitive, high-volume jobs out of UK businesses and replacing them with something that costs a fraction to run.
Systems delivered.
Two systems, both running in production.
Property-sourcing platform rebuild
Replaced an expensive third-party scraping service with a system I wrote — Python, residential proxies, real browser sessions. Structured logging then exposed that over half the daily work was being silently abandoned, and that one dead link had been burning the monthly software budget for months.
Self-optimising message selection
Instead of running an A/B test for weeks and picking a winner by hand, it reallocates sending traffic toward whatever is earning genuine human replies, and keeps adapting. It scores engagement by how fast it arrives, so inbox scanners firing in milliseconds count for nothing.
The same optimisation system also lifted positive replies by 19% on a 10,000-lead campaign at StaffCircle, where I work part-time, and by 14.2% for a separate business. Measured against equal-split baselines.
Four commitments.
No build is recommended without a case for it
If the audit says the job isn't worth automating, that's what it will say. That outcome is a success, not a failed sale.
Nothing goes live without your written approval
Every system proves itself against your own team first. You decide when it goes live, in writing.
Existing software is retained wherever possible
It's nearly always better to make what you already have work properly than to migrate everyone onto something new.
One engineer, start to finish
No account managers, no junior developer taking over after the sale. You get me, throughout.
Company details.
Trading name | ProductConsult |
Legal entity | APEX SCALE LTD |
Principal | Caius Seemann — sole engineer on every build |
Based | United Kingdom — working with UK businesses |
Data handling | UK GDPR data-processing agreement on every build |
Contact | hello@productconsult.com |
Describe the process you want to automate.
You will receive a fixed price and a build plan.
There is no charge for the initial call and no obligation. You deal directly with the engineer who builds the system.