Module 01 · Your AI Job-Search Toolkit
Audit Your Resume and LinkedIn Without Inventing Evidence
Open lesson + course map
On this lesson
Course outline
Module 1 · Your AI Job-Search Toolkit
Module 2 · Resume Engineering
Module 3 · LinkedIn as a Lead Engine
Module 4 · Interview Preparation With AI
A resume audit is a comparison between what employers repeatedly ask for and what your documents can honestly prove. AI can organize that comparison across several job posts, but it cannot know which skills you possess unless you supply evidence.
By the end, you will have one gap table built from three live job posts. Every gap will lead to either a truthful document change or a learning plan—never a made-up claim.
// concept
Choose Three Comparable Job Posts
Use three current posts for the same role family and roughly the same seniority. Compare three junior marketing roles, not an intern, a senior brand manager, and a graphic designer. Repeated requirements across comparable posts are more useful than one unusual listing.
Save each post as plain text with its URL, role title, employer, and access date. Preserve “required” versus “preferred” exactly as written.
Prepare four redacted inputs:
- your master resume;
- your LinkedIn headline, About, Experience, Projects, Education, and Skills sections;
- the full text of job posts A, B, and C; and
- a small evidence inventory: certificates, project links, dated files, or responsibilities you can verify but have not yet written down.
Remove your CNIC, address, contact details, employee IDs, references' details, and confidential employer or client data. Review the AI vendor's current data controls first. If the material is sensitive, compare the texts manually.
// concept
Use Only Two Gap Classifications
A requirement missing from the resume or LinkedIn is not automatically a missing skill. Ask which of these two situations is true:
| Classification | Evidence test | Correct action |
|---|---|---|
| Have it, didn't show it | You can point to a real project, task, course output, dated file, or consenting verifier | Add or clarify the evidence in the most relevant resume or LinkedIn section |
| Genuinely missing | You cannot yet demonstrate the skill or responsibility | Put it in a learning-and-proof plan; do not add it to the resume or LinkedIn |
AI may find a possible gap, but you make the classification. If it cannot quote evidence, treat the item as unverified. Answer specific follow-up questions; a vague “probably” is not evidence.
For a genuine gap, define a proof task. “Learn CRM” is too broad. “Complete official beginner exercises, then build a sample pipeline with fictional data” is reviewable. Until that proof exists, the resume remains unchanged.
// concept
Run the Evidence-First Audit Prompt
Paste the prompt below into your chosen AI assistant, followed by the four redacted inputs. Keep the labels and rules intact.
You are auditing a candidate's resume and LinkedIn profile against three job posts.
Your job is gap analysis, not resume invention.
INPUTS
1. REDACTED RESUME: [paste]
2. REDACTED LINKEDIN SECTIONS: [paste]
3. EVIDENCE INVENTORY: [paste]
4. JOB POST A: [role, employer, URL, access date, full text]
5. JOB POST B: [role, employer, URL, access date, full text]
6. JOB POST C: [role, employer, URL, access date, full text]
METHOD
- Extract concrete requirements using the employers' wording.
- Record which posts mention each requirement and whether each calls it required
or preferred. Do not infer importance from repetition alone.
- Quote the exact resume, LinkedIn, or evidence-inventory text that supports it.
- Omit requirements already shown clearly from the final gap table.
- For every remaining requirement, ask one factual candidate-confirmation question
if the evidence is ambiguous.
- After I answer, classify every gap as exactly one of:
"Have it, didn't show it" or "Genuinely missing."
- For "Have it, didn't show it," propose where to show the verified evidence.
- For "Genuinely missing," propose a learning-and-proof action. Never propose a
resume or LinkedIn edit for that gap.
NON-NEGOTIABLE RULES
- Do not invent skills, titles, dates, tools, duties, metrics, outcomes, or evidence.
- Do not rewrite a requirement as if the candidate has met it.
- Do not claim to predict recruiter, ATS, interview, or hiring decisions.
- Flag contradictions between the resume and LinkedIn; do not silently resolve them.
FINAL FORMAT
Requirement | Posts mentioning it | Required/preferred | Evidence found |
Classification | Action | Candidate verification neededReview each row against the original post and your evidence files. Delete any model-added interpretation you could not defend in an interview.
// worked_example
Worked Example
The following is a fictional sample, not a real candidate or hiring result.
Sample candidate evidence: Ayesha's internship notes say, “Compiled a weekly Excel summary from sales sheets submitted by four representatives for three months.” Her student-society folder contains a six-week content calendar she maintained. Her resume says only “Assisted the sales team,” while her LinkedIn About says “BBA graduate seeking marketing opportunities.” She has never used a CRM or built an analytics dashboard.
Sample extracts from three fictional posts: Post A requires Excel reporting and lists Canva as preferred. Post B requires maintaining a content calendar and lists dashboard experience as preferred. Post C requires written communication and lists CRM familiarity as preferred. The full audit prompt above is run with those excerpts plus the sample evidence.
The model's useful first pass identifies Excel reporting and content-calendar work as absent from the public documents, then asks whether Ayesha actually performed them. It also identifies CRM and dashboard experience as unsupported. After Ayesha confirms the evidence, the reviewed table becomes:
| Requirement | Posts | Evidence found | Classification | Action |
|---|---|---|---|---|
| Excel reporting | A: required | Weekly summaries from four representatives for three months | Have it, didn't show it | Replace “Assisted the sales team” with the verified scope |
| Content calendar | B: required | Six-week student-society calendar in dated folder | Have it, didn't show it | Add one project bullet and link the redacted sample if appropriate |
| Analytics dashboard | B: preferred | None | Genuinely missing | Build a fictional-data dashboard as a learning project; do not edit resume yet |
| CRM familiarity | C: preferred | None | Genuinely missing | Complete official beginner practice and create a sample pipeline first |
A bad draft-one rewrite might say: “Managed CRM-driven campaigns and analytics dashboards to improve sales performance.” Every important claim in that sentence is unsupported.
The specific fix is: “Compiled weekly Excel summaries from sales sheets submitted by four representatives during a three-month internship.” This uses only the sample evidence. CRM and dashboards stay in the learning plan until Ayesha produces proof.
// failure_cases
Failure Cases to Diagnose
6 cases to diagnose
Three unrelated roles produce a noisy list
restart with posts that share a role family and level.
The model merges required and preferred items
restore each employer's label from the saved post text.
A gap is classified from confidence rather than evidence
ask for the exact supporting quote; if none exists, verify it yourself.
A genuine gap triggers polished resume wording
remove the wording and create a learning-and-proof action instead.
Resume and LinkedIn dates or titles conflict
check original records and correct the inaccurate document; never ask AI to choose the more impressive version.
The table contains tools implied by broad phrases
“digital marketing” does not prove CRM, Canva, or analytics experience. Name only tools actually used.
// pakistan_angle
Pakistan Angle
On Rozee.pk, LinkedIn, or Indeed, save the post and URL when you apply; a mobile listing may change before an interview. During load-shedding or mobile-data interruptions, keep the post texts and gap table locally so you can verify offline.
Pakistani application forms sometimes request more personal detail than a gap audit needs. Do not paste CNIC numbers, a full residential address, salary slips, references' phone numbers, or family details into an AI chat. Replace names and contact fields with labels. Keep the original resume separately for submission through the employer's official channel.
Work from a family business, university society, local shop, or informal internship can support a requirement when described honestly. Use the role you actually held, record the scope you can prove, and ask a reference's permission before listing them. English, Urdu, or Roman-Urdu work counts only at the level your actual samples demonstrate; occasional translation does not justify claiming professional fluency.
// hands_on
Hands-On Exercise
6 steps
Build your own gap table:
Save three live, comparable job posts with URLs and access dates.
Redact your resume, LinkedIn sections, and evidence inventory.
Run the exact audit prompt and answer its factual confirmation questions.
Verify every requirement against the original post and every evidence quote against your files.
Classify each gap using only the two allowed labels.
Give every “Have it” row a truthful document action and every “Genuinely missing” row a learning-and-proof action.
// completion_rubric
Completion Rubric
6 checks — tick as you verify
// sources
Sources
3 official sources — check every claim yourself
// check_yourself
Check yourself
4 questions · answers and options are taken word-for-word from this course
1 / 4 · diagnose
Your work shows this failure mode: “The model merges required and preferred items.” What does the lesson tell you to do about it?