AI Job Description Writer
Enter a title, category, seniority and salary range — get a complete, bias-free Markdown JD with all the standard sections in seconds.
Job-creation form
AI for recruiters
Post green jobs in minutes, not afternoons.
Generate the full job description, screening questions, and a market-anchored salary range — all inline in your existing job-creation form. Toggle off for the whole org with one click.
No candidate PII to AI Internal audit log for every call Org-wide kill switch
Five AI tools
Everything you need to ship a high-quality posting fast
Every tool lives inline in the existing job-creation form. No new tabs, no separate AI page to learn.
AI JD Improver
Paste your existing JD and AI rewrites it: removes gendered language, hollow buzzwords, and vague requirements. Returns an inclusion score + list of fixes.
Edit-job formScreening Questions Generator
AI drafts 5 targeted pre-screening questions (yes/no, text, multiple-choice) tailored to the role and category. Cached 7d for the same role shape.
Question builderSalary Range Advisor
Suggests a competitive salary range in the country's local currency — INR for India, AED for UAE, EUR for Germany. Shows market position & confidence.
Salary sectionOrg-Wide Toggle
Owners flip AI on or off for the entire company from /recruit/company. Mirrors the MFA toggle. Disabled means every AI button returns 403.
Company settings
How it works
From blank form to published posting — in four moves
-
Open the new-job form
Fill title, category, seniority and location like normal.
-
Click "Generate with AI"
A full JD appears as Markdown in a preview modal — review, then accept to insert into the editor.
-
Tune screening + salary
One click each for AI questions and a country-anchored salary range. Apply pre-fills 4 form fields.
-
Publish & monitor
Every AI call is logged internally with the user, company, tokens used, and any fallback hits — for abuse monitoring and quota tracking.
Privacy & control
Owner-controlled. Audit-logged. PII-stripped.
Built for green-economy teams who care about how AI uses candidate data.
No candidate PII to AI
Candidate names, emails and phones are stripped before any AI scoring prompt. The model sees only redacted skills + role data.
Org-wide kill switch
Owners disable AI for the whole company on /recruit/company. No team member can use it while the toggle is off.
Full audit log
Every AI call is recorded internally: user, company, feature, tokens, cache hit/miss, fallback triggered. Reviewable by platform admins for abuse monitoring.
Cache where it counts
Screening questions cached 7d by (title + category + seniority). Salary ranges 24h by full input shape. Free reads when context is unchanged.
Sample request
What we send to AI for a salary suggestion
No PII. Only structured role + location signals. Cached 24h on the input shape.
Role: Senior Solar Engineer
Green category: solar-pv
Seniority: senior (8 years)
Employment type: full_time
Work mode: hybrid
Location: Mumbai, MH, India
Output currency: INR ← local currency for India
Period: gross annual
Frequently asked
What recruiters ask first
Will AI publish jobs automatically?
No. AI only generates suggestions — you always review and click publish yourself. No auto-publish, no auto-send.
Does the salary advisor return INR for Indian jobs?
Yes. The country picks the local currency via our country→currency map (INR for India, AED for UAE, EUR for Germany, etc.). USD is the fallback only when the local currency isn't in our picker.
Can my team members see what AI generated?
AI suggestions are only shown to whoever clicked the button. The final JD/questions are saved to the job once published — visible to your team via your normal job permissions.
How do I disable AI for my company?
Go to /recruit/company → Security & Settings → AI Assistants toggle. Same flow as the MFA toggle. Disabling takes effect immediately for all team members.
Does AI score candidates automatically?
Not in the current release. The scoring API exists but is not auto-invoked — we want recruiters to review applications themselves.
How reliable is the AI?
We run a primary AI model with automatic fallback to a secondary model on any failure — recruiters rarely notice an outage. Both options are tuned for fast, structured responses.