Why I built 3sixty.me — and what three AI systems thought of the result

A personal story about a broken hiring process, what I tried to do about it, and three AI systems that were asked to judge the result.

I didn’t set out to build a career app.

I’m not a careers coach, I don’t work in HR. What I do have is a close-up view of what happens when someone you care about — someone genuinely capable, with a strong track record and serious academic credentials — runs headlong into the modern job application process and comes off worse.

The person I’m thinking of had been a senior researcher in the City before going back to study. They came out of Oxford with a Master’s. By any conventional measure, they were exactly what employers should want. What followed was months of rejection letters — when they came at all — and more often just silence. Application after application, into the void. No feedback. No explanation. Just the grinding accumulation of not hearing back.

I started reading about why this happens. The picture that emerged wasn’t encouraging.

The system that never sees you

Most graduate and post-graduate applications never reach a human being. They’re screened by Applicant Tracking Systems — ATS software that filters CVs against job descriptions before a recruiter sees anything. These systems are increasingly AI-driven, running semantic matching algorithms that score applications for contextual fit, not just keyword presence. A candidate can be genuinely right for a role and score poorly simply because they haven’t framed their experience in the language the system is looking for.

The consequences for candidates go beyond the obvious frustration of not getting interviews. What I kept reading about — and what I was watching happen — was something more corrosive. Repeated rejection without feedback destroys confidence. Capable people start to doubt whether they were ever as good as they thought. They become desperate, which makes them present worse, which generates more rejection. It’s a spiral that the process itself creates, and it falls hardest on people who are early in their careers and don’t yet have the self-knowledge or evidence to push back against it.

The job market isn’t helping. Graduate vacancies in the UK fell by more than a third in the year to March 2026, according to figures published in The Times — partly economic, partly a direct consequence of AI being used to do work that graduate hires used to do. The roles are changing faster than the application process has adapted to explain what employers are actually looking for now.

What I thought I could do about it

I couldn’t do anything about the number of jobs available. But I kept thinking: surely it’s possible to do something about the process. I’ve spent most of my career working in and advising smaller businesses — I chair the Wales Management Council and recently co-founded Business Hwb, a partnership to support better small business management in Wales — and the gap between what capable people can do and how they present themselves is something I’ve watched play out on both sides of the hiring table. To give candidates a fighting chance of getting through to a human being felt worth attempting.

That’s what 3sixty.me is.

The starting point is peer feedback — structured, specific, gathered through the app using a recognised strengths framework. People who know you well are asked to identify your strengths and share a real example of when they saw one in action. That evidence is mapped against your own self-assessment using the Johari Window model, showing you where you and others agree and — more interestingly — where you and they don’t. The strengths you didn’t know you had. The blind spots that turn out to be assets.

From there, an AI-led conversation takes the peer evidence and connects it to a specific role — reading the job description, mapping your strengths against what the employer is actually asking for, identifying gaps honestly, and building a strengths-backed case you can use in applications, interviews, and cover letters. The goal isn’t to help candidates claim things they can’t evidence. It’s to help them present what they can genuinely demonstrate, in language that works for both the algorithm and the human reading behind it.

What three AI systems made of it

I wanted to test whether it actually does that. So I took a real 3sixty report — generated for a graduate applying for a Sales Development Representative role — and put it in front of three AI systems separately. Gemini, ChatGPT and Claude. Each working independently, each given the same input. The question was simple: would using this advice strengthen the application and help get the candidate to the top of the list? I also thought there was something ironically delicious about asking 3 AI platforms to mark AI’s own homework giving people a way to get around AI constructed barriers.

All three came back with broadly the same answer.

GEMINI

Gemini identified what it called the single biggest structural flaw in early-career applications and explained directly how the report addresses it.

“The single biggest flaw in early-career applications is unsubstantiated self-praise. The report completely rewires that dynamic by using third-party validation. When a candidate uses specific names, data points, and peer backings, the recruiter’s scepticism drops. It transforms subjective claims into objective facts.”

CHATGPT

ChatGPT focused on what the evidence actually does in practice.

“It does not just say ‘good communicator’ or ‘team player.’ It gives evidence, context, outcomes, and direct relevance to the job. That is exactly what hiring managers and recruiters look for.”

CLAUDE

Claude picked up on something I hadn’t anticipated — the value of the blind spot analysis.

“The blind spot awareness section signals self-awareness, which is genuinely rare at graduate level and something sales managers value.”

What the agreement means

None of them said it guarantees the top spot. All three were careful to qualify their conclusions — the competition matters, experience gaps need honest bridging, the language needs to sound natural rather than scripted. That felt like a valid answer. A product that claims to guarantee outcomes in a competitive job market isn’t being straight with you.

What I found striking was that three systems — trained differently, with different architectures and different tendencies — reached pretty much the same conclusion when given the same question and input. That’s not a marketing claim. It’s three independent reads of a real document arriving at the same place.

What it can and can’t do

3sixty.me won’t solve the graduate jobs crisis. It won’t create roles that AI has displaced or reverse the structural pressures the market is under. What it might do is give candidates a meaningfully better chance of getting through to a human being — with specific, peer-evidenced proof of what they’re genuinely good at, framed in a way that works for both the systems and the people on the other side of the process.

That felt worth building.

If you’re a graduate or post-graduate navigating this market right now — or if you know someone who is — I’d genuinely like to hear what you make of it. The product is at 3sixty.me.
Step 1 - Discover; which gives you your peer assessed strengths report will always be free.
Step 2 - Deploy; which gives you detailed reports to help you position yourself for a job application, promotion or simply to get better at what you do is free for a limited period.