TL;DR
A 60-second “tell me about yourself” is roughly 130-150 spoken words (interview pace runs 120-150 words per minute; 140 wpm reads as the most credible). Use the recruiter-standard Present → Past → Future arc: one line on who you are now, two highlights anchored in real numbers, one line on why this role. AI is good at compressing that into 60 seconds and varying the register; it cannot pick your two highlights, so you feed it the job description and your metrics. The prompt below does the structuring. Use GPT-5.5 for tight, metric-forward drafts and Claude Sonnet 4.6 / Opus 4.7 when you want a more natural spoken cadence.
The task
First round, and the first question is “So, tell me about yourself.” You have about 60 seconds before the interviewer forms an impression that frames the rest of the conversation. Research on hiring decisions puts that impression inside the first minute, so the opening is not a warm-up. The common failure is reciting your CV in chronological order and ending with “and that brings me here today.” The job is a self-intro that earns specific follow-ups, not generic ones, and that hooks back to this role.
When AI helps, and when it does not
AI is excellent at structuring a 60-second arc, tightening adjectives, cutting to the word count, and switching between a technical and a non-technical register. It is bad at choosing your two highlights; that decision is yours, based on the job description. Feed AI the JD and your real metrics, or you get a generic intro that fits everyone and convinces no one. A 2026 caveat: leading models still occasionally inflate a number or invent a “result” you never claimed, so verify every figure in the draft against what actually happened.
What to feed the AI
- Your background in one sentence (current role, years, domain)
- The full JD you are interviewing for (paste it; do not summarize)
- 2 highlight stories with concrete metrics (users, revenue, latency, time saved, scope, team size)
- One specific thing about this role or company that excites you, researched, not generic
- Banned tics: “passionate about,” “self-starter,” “fast learner,” “team player”
- Whether the panel is technical or non-technical (changes vocabulary and how much jargon survives)
Copy-ready prompt
Write a 60-second interview self-introduction.
Background (one sentence): [line]
JD I am interviewing for: [paste full text]
Highlight 1 (with metrics): [line]
Highlight 2 (with metrics): [line]
One specific reason this role excites me: [researched, specific]
Banned tics: [list]
Panel type: [technical / non-technical]
Structure (Present -> Past -> Future):
- One sentence background: current state + adjacent expertise
- Highlight 1: ~15 seconds, anchored in a metric
- Highlight 2: ~15 seconds, anchored in a metric, a different skill from highlight 1
- One sentence on why this role specifically
Constraints:
- Total 130-150 words (about 60 seconds at interview pace)
- Each highlight must earn a specific follow-up: a number, a trade-off, or a decision
- Last sentence must reference the JD specifically: name a product, a metric, or a stated value
- No banned tics, and strip every adjective that is not load-bearing
- Plain spoken sentences I can say out loud, not written-essay phrasingVariant for senior roles: “Same structure, but assume the panel knows my level. Lead with one strategic contribution instead of two tactical wins, and frame the ‘why this role’ around scope or ownership.”
Which model to use
Both major assistants handle this well on their free tiers (as of June 2026). The difference is texture, not capability.
| Tool | Best for this task | Free tier | Notes |
|---|---|---|---|
| GPT-5.5 (ChatGPT) | Tight, metric-forward drafts that enforce structure | Yes (tight limits) | Strongest at hitting the word count and pushing every claim toward a number |
| Claude Sonnet 4.6 / Opus 4.7 | Natural spoken cadence, mock follow-ups | Yes (Sonnet 4.6, limited) | Reads more like talking; Opus 4.7 on Pro ($20/mo) for the back-and-forth practice |
| Gemini 3.1 Pro | Drafting if you already live in Google Workspace | Via Google AI ($0 limited) | Fine for a first pass; tends to run long, so enforce the word cap |
Practical move: draft with GPT-5.5 for the structure, then paste into Claude and ask it to “say this out loud, make it sound spoken, not written.” You get the discipline of one and the warmth of the other.
Recommended output structure
Three short spoken paragraphs, 130-150 words total. A timing note at the end (your reading speed in seconds). Two alternate openers in case nerves blank the first line.
How to check the output is usable
- Time it out loud at interview pace, then trim to 55-65 seconds
- Each highlight has a number, a trade-off, or a decision the interviewer can ask about
- The “why this role” line could not be pasted into another company’s interview
- No banned tic survived, and the adjective count is near zero
- The intro invites the follow-ups the interviewer would have asked anyway, so you steer the next five minutes
Common mistakes
- Listing every job you’ve held: the panel already has your resume
- No metrics: a claim without a number is filler the interviewer cannot probe
- Generic “why this role”: interviewers detect copy-paste in about five seconds
- Adjective stack (“driven, passionate, collaborative”): strip every adjective on the second pass
- Running past 75 seconds: cut adjectives and connective phrases, never the stories
- Reciting it word-for-word: sounds rehearsed and triggers skepticism, so memorize the beats, not the script
- Trusting AI’s numbers: it can round 30% up to 40% or invent a “result,” so check each figure
FAQ
- Can I use one intro for every interview? The background line can stay. The two highlights and the “why this role” line should change for each company, because those are what the AI tailors from the JD.
- Should I open with a hook? Only if it genuinely fits the story. Most hooks feel forced, so default to a direct, confident first line.
- What if they cut me off? Plan a “skip to highlight 2” path and rehearse the ending without highlight 1, so you can still land the “why this role” close.
- Is 160 words too many? Slightly. At 140 wpm, 160 words runs close to 70 seconds. Target 130-150 words; it leaves room for the natural pauses you’ll add when nervous.
- Can AI practice the delivery with me? Yes. ChatGPT and Claude both have voice modes, so you can speak your intro and ask the model to time it, flag filler words, and fire one follow-up per highlight.
Related
- STAR interview answers: the body of the interview after the intro
- STAR follow-up questions: prep for the questions your highlights invite
- Mock interview AI: practice the full opening out loud
- AI resume writing: the resume this intro condenses
- Job description analysis: what to mine from the JD
- Behavioural story mining prompts: surface the highlight stories
Tags: #AI writing #Job search #Workflow