For engineers who want an honest answer
Know exactly where you stand
before you apply.
Upload your résumé and a job description. Reflect analyses your skills, experience, domain fit, and evidence depth — then tells you honestly whether to apply, and exactly what to fix if you should.
Built for software engineering and technical roles.
Watch it reason,
not match.
This is Reflect mid-analysis — reading each line for what is actually evidenced, the way a senior engineer would, before it reaches any verdict.
Reflect analysing resume.pdf against Senior Backend Engineer.
- Résumé: “Built and shipped production APIs in Python and PostgreSQL for three years.” — Evidenced — production context, not coursework. (Evidence Depth)
- Résumé: “Led migration to containerised deployments.” — Claimed. No supporting detail found. (Evidence Depth)
- Job description: “Senior engineer, 5+ years, distributed backend systems.” — Role expects five years at senior scope. (Experience)
- Résumé: “Owned on-call rotation for the core payments service.” — Production ownership — evidenced, directly relevant. (Domain Fit)
Reasoning complete — your verdict is ready.
Every tool has an opinion.
Few have evidence.
The same candidate, the same role — read two different ways. Throw the switch.
The same candidate and role, read two ways. Other tools report: Match score: 87%, Top 15% of applicants. Keyword match: 4 / 4, Every required keyword found in your résumé. Verdict: Apply now, You’re exactly what they’re looking for. Tone: Encouraging, Every résumé looks like a great fit.
Reflect reports: Overall fit: 64%. Weighted across four dimensions — with the evidence behind every point, and every gap. Evidence check: 3 evidenced. Three are backed by real project evidence. One is only claimed. Verdict: Apply after addressing gaps. One gap — Kubernetes — stands in the way. A few weeks of work closes it. Tone: Honest. Calibrated to the evidence, not to encouragement. It says so when the case is thin.
A lower number with the reasoning behind it is worth more than a high one without — it’s also why you can trust the verdict when it’s good.
Sincerity isn’t the same as rigour.
Here’s the rigour.
Four properties of how Reflect scores — not steps to watch, but guarantees you can rely on.
Not one number — four.
Skill, experience, domain, and evidence — each weighted independently, then combined.
A high score doesn’t guarantee a good verdict.
A structural gap in what the role requires can override it completely.
Run it twice, get the same answer.
Same résumé, same role — identical score, tier, and decision, every time.
When the evidence is thin, Reflect says so.
Rather than rounding up to sound more confident than the case actually is.
That’s the discipline behind every verdict.
Here's what one actually looks like.
An excerpt from an example Reflect report — the real document, for an invented candidate. Scroll to inspect it.
Example report
One verdict. Four dimensions behind it. Nothing softened, nothing hidden.
An example Reflect report: Stretch match, apply after addressing gaps, 64% overall fit. Evidence across four weighted dimensions — skill, experience, domain, and evidence strength — supports the verdict; a specific gap (Kubernetes) is named rather than left vague. The report also names risks worth accepting and signals inferred from the résumé beyond the verified evidence.
Before you ask.
Is this just an LLM guessing?
No — the score, tier, and decision are computed deterministically across four dimensions. The only thing an LLM writes is the explanation alongside them, never the number itself.
Would I get the same result twice?
Yes, for the score, tier, and decision. Only the wording of the explanation can vary between runs.
What happens to my résumé?
Your file is never stored — it's processed in memory and discarded immediately. The text extracted from it is kept for 90 days, and you can ask for it to be deleted sooner; our Privacy & Trust page sets out exactly how.