About RecastCV
RecastCV exists because we were frustrated with what every other AI resume tool was doing. Not with the idea of AI-assisted job applications — that part made sense. The frustration was with how those tools handled the most important constraint: the truth.
Where this started
The team behind RecastCV spent years on both sides of the hiring table. We have written CVs for ourselves, reviewed CVs from candidates, and sat through enough screening calls to understand what actually matters. When large language models became capable enough to help with resume writing, we tried every tool we could find. The results were consistently disappointing — not because the writing was bad, but because the writing was fictional.
Tools would invent metrics. They would attribute outcomes the candidate had nothing to do with. They would lift skills from the job description and embed them in the CV as though the applicant had always had them. The output looked polished. It was also, in many cases, a lie.
That is a problem for the candidate, not just the recruiter. When someone interviews for a role with a CV that overstates their experience, the interview falls apart. The mismatch between what the CV promised and what the candidate can actually demonstrate is immediately obvious. We have watched that happen. It is not a good outcome for anyone.
The grounded tailoring thesis
Our answer was a constraint we call grounded tailoring. The idea is simple: the AI can only use material that already exists in your master CV and your project library. It can reframe. It can emphasise. It can reorder and rephrase to match the language of the job description. What it cannot do is invent.
In practice this means you upload your real work history once — your master CV, plus any project details you want the system to draw from. When you apply for a role, you drop in the job description URL. RecastCV reads the JD, identifies the keywords and requirements that matter, and rewrites each bullet of your CV to surface the closest matching experience from your actual background. Every output is traceable back to something you genuinely did.
The result is a CV that is both more relevant and more defensible. You can walk into an interview and answer every question about what is on the page, because everything on the page is real. The ATS score tends to improve as well — not because we stuffed keywords in, but because relevant experience is now framed in the language the hiring system is looking for.
"The best tailored CV is the one where every sentence is true — just expressed in the language the role is hiring for."
Building for both CV and resume markets
The UK and US job markets use different terminology and different conventions. A CV in the UK is typically two pages with a personal statement; a resume in the US is often one page without one. Recruiters in different markets weight different sections differently. Most AI resume tools are built by and for the US market and treat everything else as an afterthought.
RecastCV was built to handle both. The tailoring engine understands the structural differences between a UK CV and a US resume. You can apply for roles in London and San Francisco with the same master CV, and the output will be appropriately formatted for each.
The iOS app
RecastCV is also available as a native iOS app. The iOS app and the web app share the same back-end infrastructure — the same Supabase project, the same tailoring engine, the same project library. If you upload your master CV on the web, it is available on mobile. If you track an application on iOS, it appears in the web dashboard.
We built the iOS app because job searching happens everywhere — on the commute, between meetings, late at night. Being able to spot a role, pull up the JD, and generate a tailored CV in under a minute from your phone is genuinely useful. The credit system is shared across both platforms, so there is no friction switching between them.
What we are building toward
The application tracker, the interview prep module, and the ATS checker are all part of the same thesis: job searching is a process, not a moment. Most tools optimise for the CV in isolation. We are trying to optimise for the whole loop — from finding a role, to tailoring your application, to preparing for the call, to tracking where every application stands.
We are a small team. We move deliberately. We do not ship things that do not work, and we do not add features that require you to trust the AI more than you trust yourself. If you have feedback, particularly if you have found something that felt wrong or dishonest in the output, we want to hear it. Contact us at ai@sumletter.com.
The technology
RecastCV is built on Supabase for data storage and authentication, Anthropic Claude for the tailoring engine, and Stripe for payment processing. We use a credit-based model — each tailored CV costs one credit. There is no subscription, no auto-renewal, no lock-in. You buy credits when you need them and use them when you apply.
The tailoring pipeline sends your master CV and the job description to Anthropic's API to generate the rewrite. Your data is stored in Supabase's EU region. For full details on how we handle your data, see our privacy policy and security page.
Start tailoring your CV today.
Three free credits on sign-up. No credit card required.