Resume tailoring.One API call away.
Refleum AI accepts a resume and a job description, and returns a tailored resume, cover letter, and outreach message — grounded strictly in what the candidate already has.
POST /api/v1/resumes/:resumeId/tailor → 201 in < 30 s
Powering job-search products at
Built for truthful tailoring.
Every change is grounded in what the candidate actually has. The pipeline never fabricates — it only refines.
Diff-Based Pipeline
The LLM produces a targeted list of changes — not a full rewrite. Every diff is verified against the original before it applies. Rejected diffs are counted and returned.
Multi-Pass Refinement
After tailoring: keyword injection from the master resume, AI-phrase removal (em-dashes, "spearheaded", "leveraged"), and alignment validation — all automatic, all reported inrefinement_stats.
Org-Level Isolation
Every resume, cover letter, and usage record belongs to exactly one organisation. Row-level isolation on every table. Cross-tenant access is architecturally impossible.
Truthfulness Constraint
The alignment validator automatically removes fabricated skills, certifications, and employers before the response reaches you. No invented metrics. No hallucinated employers.
Cover Letter & Outreach
Request a cover letter and LinkedIn outreach message in the same tailor call. Generated in parallel, stored separately, regeneratable any time.
PDF Export
A4 or Letter, multiple templates, custom margins and fonts — rendered by headless Chromium.
5 Languages
EN · ES · ZH · JA · PT — specify per request.
What happens in a single tailor call
- 01
Keyword extraction
Required skills, preferred skills, responsibilities, and seniority parsed from the JD.
- 02
Diff generation
LLM outputs a targeted change list — paths, actions, original text, and reasons.
- 03
Safety nets
personalInfo locked, dates restored, dropped skills re-appended, custom sections protected.
- 04
Multi-pass refinement
Keyword injection → AI-phrase removal → alignment validation. Stats returned.
Developer Resources
Everything you need to integrate.
Simple, metered pricing.
Pay for tailor calls. PDF exports and reads are always free.
Perfect for hobbyists and side projects.
- 1 request per minute
- 3 team members
- Community support
- Basic analytics
Perfect for growing teams and startups.
- 20 requests per minute
- 10 team members
- Priority email support
- Standard analytics
The professional choice for scaling startups.
- 60 requests per minute
- 100 team members
- Priority support
- Advanced monitoring
- Custom domains
Bespoke solutions for large organizations.
- 1000+ requests per minute
- Unlimited team members
- Dedicated support
- SLA guarantees
- Custom integrations
Tailor calls are metered per request — billed even when the pipeline fails, because every call triggers LLM API costs. Reads and PDF exports are always included. Billed via Polar.sh.
Ready to transform your dev workflow?
Join thousands of teams shipping faster with a production-ready foundation.