Manchester studio · Bespoke internal tools
What we do
Four common shapes the work takes. What we build is bespoke to the job your team does by hand.
AI content QA & compliance
Upload your guidelines once and every draft gets a first-pass review against them, checking references, claims and house style before a person even looks.
Selected work
Real tools, running in real businesses. Different industries, one job: take the manual work off the team.

A pricing system that manages luxury Edinburgh short-let rates twelve months ahead, plus an AI phone assistant that turns missed calls into instant replies and concise summaries.

An AI first-pass reviewer that checks promotional manuscripts against client guidelines, references and claims before they reach an editor, reducing repetitive editorial QA.

A custom admin dashboard that runs newsletters, subscribers, blog and inbound email for three brands, from one place.

One app that runs a jewelry studio's customer catalog and its staff point of sale, with live inventory that can't be oversold.

One portal for students, parents, mentors and admins that runs billing, scheduling, mentor payouts and reminders for a premed mentorship program.
How we work
No long proposals. We look at the actual work, build the piece that removes it, and keep it running.
A free 30-minute workflow audit. We find the task quietly eating your team's week, and what it costs you in hours.
A one-time build. We make the tool that handles that job for you, wired into the software you already use.
A monthly maintenance and support retainer. Hosting, running costs, monitoring and updates, so it keeps working as the business grows.
How we handle your data
These tools sit on top of your business. For most of our clients, how the data is handled matters as much as what the software does, so we build with that in mind from the first line.
We don't use your business data to train models. Where we use third-party AI providers, we use their business or API products with model training disabled, and document those providers with you.
A Data Processing Agreement is in place for every project where we handle data on your behalf, with subprocessors documented.
Clear rules for retention, deletion and access are part of the build, not an afterthought bolted on later.
Encryption in transit and at rest, access limited to what each part of the system actually needs.
Your tool runs on infrastructure that is already certified. Supabase is SOC 2 Type 2 and ISO 27001 certified, and the providers we build on operate under signed data processing agreements with model-training on your data switched off.
supabase · vercel · next.js · openai · anthropic · stripe · twilio · resend
The studio
Third Space Labs is a small studio. We work directly with the people who run the business, no account managers and no four-week discovery phase before anything gets built.
We start from how your team actually works, pick the one job that costs the most time, and ship a tool that handles it. You talk to the person writing the code the whole way through.
FAQ
It depends on the job, but most work is a one-time build fee plus a monthly maintenance and support retainer. The free audit gives you a real number for your specific workflow before you commit to anything.
Most tools go from audit to something you can use in a few weeks, not months. We start with the one job that costs you the most time and ship that first, then build from there.
The bespoke tool we build for your business is yours to use, without per-seat licensing or lock-in, and your contract sets out ownership clearly before we start. If you ever stop working with us, it keeps running.
It stays yours. We don't use your business data to train models, and where we use third-party AI providers we use their business or API products with model training disabled. There is a data processing agreement in place for every project.
Yes. The monthly retainer covers maintenance, monitoring, model and dependency updates, and fixing anything that breaks, so the tool keeps working as the business grows.
Your choice, and we settle it when we agree the work. You can own the accounts and pay the providers directly, which keeps your data in your own org, or we can run it in ours and pass the usage through. Either way, the running cost is just what the providers charge.
Almost always. If you want proof before a full build, we can scope and build a small working prototype of the core flow for a fixed fee, yours to keep either way.
Start here
Thirty minutes. Show us the most tedious thing your team does on a computer, and we’ll tell you what we can automate and what it would save.
Book a free auditNo commitment. Prefer proof first? We can build a working prototype of the core flow for a fixed fee, yours to keep either way.
Or email hello@trdspc.com