Founder · live
Vital
Every nurse and carer, cleared before the shift.
Vital keeps every care worker’s DBS check, right to work and training in date, automatically.
- Role
- Solo founder: I designed, built and sell it
- Company
- Vital Credentials Ltd, London
- Markets
- UK (main) and South Africa
See it working
- BeforeOrigin Studios
- 2025Idea, after school exams
- 2026Vital built
- 2026Technical review
- Oct 2026Outreach and first demo
- NextPhase 2
The problem
Hundreds of workers, thousands of expiry dates, one spreadsheet.
UK care agencies must prove every worker is cleared before each shift: a DBS check, right to work, professional registration (such as an NMC PIN for nurses) and mandatory training. Most track this in spreadsheets and chase workers by email and phone. Credentials expire on different dates, and an agency can have hundreds of workers with five to ten credentials each.
What’s at stake
An expired DBS check or registration means an unverified carer on shift, which puts vulnerable people at risk. Failing a CQC inspection can mean penalties, lost contracts and a damaged reputation.
Why existing tools fall short
Similar products always miss something: worker self-upload, document checking, flexible reminders or inspection-ready reports. Agencies end up using several tools plus a spreadsheet.
Where it came from
“My first business taught me what a hard market looks like. So I went looking for the opposite: a large market, recurring revenue, and businesses with a problem they’re required by law to solve.”
Origin Studios, my first business, published fast-paced D&D one-shots and taught me cold email and influencer outreach. But it was a small market, every sale was a one-off I had to keep finding, and the community resisted paying. After my exams I set four criteria for the next one:
- A large market
- Recurring revenue
- Businesses that want to buy
- A problem AI can solve
Credential compliance met all four: every agency has to do it by law, it never ends, and it’s mostly paperwork.
The product
- 01
Import
Upload your existing worker spreadsheet and Vital builds every worker profile. It recognises known formats, such as the South African FSCA register, and maps the columns for you.
- 02
Collect
Each worker gets a personal, secure upload link and sends documents from their phone, without creating an account.
- 03
Check
AI reads each document, extracts the name, licence number, issuing body and expiry date, and compares them with what’s expected and with official registers the agency uploads. Uncertain results go to a person for review.
- 04
Stay compliant
Vital emails workers before anything expires, and the dashboard shows who is cleared to work today.
AI document checks
A pass, uncertain or fail result with a confidence score. A person always makes the final call.
Official register matching
Agencies upload any regulator’s list and documents are checked against it. Not tied to one country.
Reminder schedules
Reminder timing set separately for each credential type.
Live compliance dashboard
Who is cleared, who is expiring and who is missing documents.
Inspection-ready reports
PDF and CSV audit reports in one click.
Placement check API
Rostering and booking systems ask “can this worker be placed?” and get an instant answer.
The numbers
- Total market
- ~£58m a year
- Targeted now
- £31m+ a year
- 3-year target
- ~£310k a year
- Outreach
- ~360 agencies
- Countries
- 2
- External review
- platform45
1.6m adult social care workers in England, at about £3 a worker a month
About 13,000 home care and healthcare staffing agencies
1% of the targeted market, about 130 agencies
Contacted in October 2026
UK and South Africa
Technical due diligence; main issues fixed
Room to grow: these figures leave out Scotland, Wales, Northern Ireland, NHS staffing, South Africa and other regulated industries. Sources: Skills for Care (2025), IBISWorld, Homecare.co.uk.
Early traction
From 360 cold emails to a pilot discussion with a UK nursing agency in under three weeks.
- About 360 UK care agencies contacted by cold email in October 2026.
- First demo booked.
- Meetings with a UK nursing agency on 21 and 23 October, with a pitch deck made for them and a pilot under discussion.
- Conversations in South Africa: Placeholder.
Pre-revenue, with live pilot conversations from my first outreach campaign.
Finding customers
A sales process, built like a product.
I used AI to research UK care agencies (CQC registration, locations, services) and write a personalised first line for each email, then tracked every lead through each stage: sent, follow-up, reply, demo, trial, customer.
Five versions, tested
- The original email
- An offer-led call to action
- A hook based on a real compliance risk
- A question as the subject line
- A short email with no price
Done by the rules
Only UK limited companies are emailed, in line with UK email marketing law (PECR). Every email has an unsubscribe line, and anyone who opts out is never emailed again.
Built alone, reviewed outside
About 37,000 lines of TypeScript, by one person.
Next.js, Postgres (Neon), Clerk for logins, Stripe for billing, Cloudflare R2 for files, Resend for email and Claude for document checks. Each agency’s data is kept separate twice, in the app and in the database itself with row-level security, so a bug in the app still can’t leak one agency’s data to another. It handles sensitive personal data under UK and South African data protection law, with retention settings and automatic deletion when an agency closes.
I hired platform45, a software consultancy, to review the code, security and operations as an investor would. The result was 3/5 Amber with a list of findings, and I answered every one in a public log: what was fixed, where, and what’s left.
“Vital decides whether each worker is compliant in two places: in the app when something changes, and in a nightly database job across every credential. The review found the two had drifted. The nightly job could mark a rejected document as compliant, which for a care agency means placing an unverified nurse. I fixed the logic and built a test that runs all 288 status combinations through both and fails if they ever disagree.”
- Failed payment events from Stripe are now retried and checked every night.
- Patched all 19 known security issues in third-party packages, 11 of them high severity.
- Added automatic security scanning (CodeQL and a secret scanner), which caught a shell injection.
- Every change now runs type checks, tests against a real database and a minimum test coverage.
Phase 2
From a tool for one agency to a network for the sector.
Portable credentials
A worker is checked once and carries that profile to every agency they work for.
Compliance officers
Independent officers join and take on checking and audit work for agencies without in-house staff.
Matching
Agencies find verified workers and officers in one place.
Each new agency brings workers, and each worker on the platform makes Vital more valuable to the next agency. The first product already collects the data the network needs.
- Business model
- Placeholder
- When
- Placeholder
What I’ve learned
- 01
Pick a market that wants to buy.
At Origin Studios every sale was a fight. Care agencies are required by law to solve this problem, so they welcome a fix.
- 02
With AI, building is fast. Knowing what to build is the hard part.
Competitors were always missing something, so talking to agencies told me exactly what to build next.
- 03
Ask for criticism early.
Paying for an outside review before anyone required it found a serious bug before it reached a customer.
- 04
Sales can be built like a product.
Measuring and testing outreach turned cold emailing into something I can repeat.
- 05
Where you sell matters as much as what you sell.
Pricing conversations in both the UK and South Africa showed me how much it matters to earn in a strong currency.
What’s next
- First paying pilots
- Turn demos into customers
- Launch in South Africa
- Plan Phase 2