iFormat Branding
An enterprise platform empowering professionals and businesses with AI resume optimization, automated job matching, brand equity coaching, and secure Stripe billing.
- >Dual-persona RBAC (Professionals & Companies) with OTP + Google OAuth auth pipeline
- >OpenAI GPT-4o powered Resume Builder, Job Matching, and Cover Letter engine
- >Stripe subscription webhooks with entitlement provisioning and AWS S3 document presigning
Project Vision & Meaning
iFormat was conceived to bridge the gap between ambitious professionals and corporate recruitment. Traditional career tools offer static template builders that fail modern Applicant Tracking Systems (ATS), while companies struggle to assess genuine candidate competencies. iFormat combines generative AI career intelligence with an executive branding suite, offering interactive CV generation, AI-powered job matching, structured coaching plans, and flexible company sponsorship tiers in a unified ecosystem.
Real-World Problems Solved
Architected and deployed a full-stack platform featuring a Next.js 16 frontend and a robust Express.js / TypeScript microservice backend. Integrated OpenAI for real-time resume parsing, keyword scoring, and cover letter crafting. Implemented Stripe multi-tier subscriptions with webhook verification, S3 presigned document uploads, and automated OTP verification to protect user identities and payment streams.
Challenges & Solutions
Multi-Turn OpenAI Resume Parsing and ATS Keyword Extraction
Problem Context:Parsing user career histories and unstructured resume uploads into structured JSON models often exceeded token limits, introduced hallucinated job entries, or suffered long gateway timeouts that froze the frontend.
Designed a chunked AI processing pipeline with structured JSON schema outputs using Zod and OpenAI's function/tool calling. Implemented prompt guardrails and background processing with optimistic client UI updates, generating ATS-optimized bullet points, skill match percentages, and formatting scores in under 2 seconds.
Stripe Webhook Idempotency and Dynamic Tier Entitlements
Problem Context:Handling multi-tier subscriptions (Starter, Professional, Executive, Enterprise) with Stripe webhooks across concurrent upgrades, downgrades, and payment failures risk creating race conditions or granting unauthorized AI generation quotas.
Implemented an idempotent webhook processor in Express that records Stripe event IDs in a processedEvents table with PostgreSQL row locking. Decoupled billing status from feature entitlements using a centralized EntitlementService, ensuring AI credits and resume export permissions instantly reflect payment confirmations with zero quota duplication.
Secure S3 Presigned Uploads with Role-Based Document Access
Problem Context:Users upload sensitive documents (resumes, company verification certificates, media portfolios). Routing raw binary uploads through the Express API server wasted server memory and bandwidth and exposed private documents to unauthenticated scraping.
Implemented direct-to-S3 uploads via AWS SDK v3 presigned URLs with strict Content-Type and size restrictions. Private documents are stored under encrypted, user-scoped S3 buckets and accessed only via short-lived temporary signed URLs generated upon verified JWT/session authentication, completely offloading file transit from the API server.