AI-Powered Resume Builder | FacileCV

Client: TalentCraft SAS, Paris, France

Architected and built FacileCV.ai as a production-grade SaaS platform using a Next.js frontend with Deno microservices backend. Features 50+ ATS-optimized templates, dual AI providers, full admin control, Grafana/Prometheus observability, and a scalable monetization system.

50+

ATS Templates

<10min

CV Generation

AI Providers

The Challenge

TalentCraft needed a resume builder that could compete in a market dominated by subscription-heavy, monolithic tools with basic AI and no operational visibility. The platform had to support multi-language content generation, handle unpredictable AI API costs, and scale reliably without becoming a maintenance burden. The client required full transparency into system health, AI expenses, and user behavior from day one.

Our Solution

We designed FacileCV around a clean microservices architecture with strict separation of concerns. The Next.js frontend handles the user experience while independent Deno microservices own AI orchestration, PDF generation, payment processing, and background jobs. Each service is independently deployable, observable, and scalable. The result is a platform that is robust in production and easy to maintain as the product evolves.

Architecture & Separation of Concerns

The system follows a clean microservices architecture. The Next.js application serves the frontend and handles SSR, routing, and authentication. Behind it, purpose-built Deno microservices handle distinct domains: an AI service manages prompt orchestration and provider switching between OpenAI and Google Gemini, a PDF service handles template rendering and export via AWS S3, a payments service wraps Stripe for subscription lifecycle management, and a notification service manages transactional emails via configurable SMTP. Each service has its own responsibility boundary, communicates over well-defined APIs, and can be deployed, scaled, and updated independently without affecting the rest of the system.

Observability with Grafana & Prometheus

Every microservice exposes Prometheus metrics endpoints tracking request latency, error rates, throughput, and service-specific business metrics (AI tokens consumed, PDFs generated, payment events processed). Grafana dashboards provide real-time visibility into system health, service-level SLAs, and resource utilization. Distributed tracing spans the full request lifecycle from the Next.js frontend through each downstream Deno service, making it straightforward to identify bottlenecks and debug failures across service boundaries. Alerting rules trigger on latency spikes, error rate thresholds, and AI cost anomalies, ensuring issues are caught before they impact users.

AI Content Engine

The AI microservice orchestrates a dual-provider system powered by OpenAI and Google Gemini. It handles CV section generation, achievement rewriting, cover letter composition, and job-description keyword optimization. The service abstracts provider differences behind a unified API, enabling hot-switching between providers based on cost, latency, or quality without any frontend changes. System prompts are versioned and configurable from the admin panel. Every AI request is metered with token counts, latency, and cost-per-call tracked in Prometheus for real-time expense reconciliation against provider billing.

50+ ATS-Optimized Templates

Over 50 professional templates across six categories: Minimal, Modern, Creative, Executive, Professional, and Tech. Every template is engineered for ATS compatibility with proper heading hierarchy, clean parsing structure, and machine-readable PDF output. The PDF microservice handles rendering, ensuring template logic is decoupled from the main application. Users can fully customize colors, fonts, and layout while maintaining ATS compliance.

Multi-Language Support

The entire platform supports full English and French localization across the UI, AI-generated content, templates, and exported documents. Locale-aware routing, translated UI strings, and language-specific AI prompts are built into the architecture. The system is designed for easy expansion to additional languages without structural changes.

User Platform & Document Management

Users manage multiple CV profiles (tiered by plan), generate unlimited cover letters, and export watermark-free PDFs. The platform supports social login via LinkedIn and Google alongside email/password auth with two-factor authentication. A step-by-step builder with real-time preview guides users through personal info, experience, education, skills, and cover letter sections. Draft and published states let users iterate before exporting.

Admin Panel & Configuration

A dedicated admin dashboard provides full operational control: user management (view, edit, suspend, delete), subscription tracking, coupon creation and distribution, contact query management, and email template configuration with custom SMTP settings. AI providers are configurable from the panel including model selection, system prompt management, and cost thresholds. Storage (AWS S3) and payment (Stripe) credentials are admin-configurable, meaning infrastructure changes require zero code deployments.

Advanced Analytics & AI Expense Tracking

The analytics layer pulls from both Prometheus metrics and application-level event logs to provide daily, weekly, and monthly breakdowns of AI usage across providers. Cost-per-request tracking enables real-time reconciliation against OpenAI and Gemini invoices. Grafana dashboards surface top users by AI consumption, provider performance comparisons, and cost trend analysis. User activity logs, export statistics, and subscription conversion funnels provide the business intelligence needed to make data-driven product decisions.

Monetization & Subscription System

The payments microservice implements a three-tier one-time payment model (Free, Solo at 9.90 EUR, Candidates at 19.90 EUR) via Stripe, deliberately avoiding subscription fatigue. Each tier controls CV profile limits, AI action quotas, watermark removal, and access duration. The service handles plan expiration with graceful downgrade (saved CVs are never deleted), coupon/promo code redemption, billing history, invoice generation, and webhook-driven state synchronization with the main application.

Security, Compliance & Infrastructure

The platform is built to meet international security standards and GDPR requirements end-to-end. The Next.js frontend is deployed to Vercel with Cloudflare for DDoS protection and edge caching. Deno microservices run in containerized environments with independent scaling. Authentication uses secure password hashing (bcrypt), JWT with JWKS key rotation managed from admin, and optional 2FA. All inter-service communication and external traffic is encrypted via TLS 1.3. PDF delivery uses AWS S3 with time-limited signed URLs. Payment processing is PCI DSS compliant through Stripe. The platform enforces GDPR across every layer: explicit user consent collection, no advertising cookies or third-party tracking, full data portability and export, right to erasure with user-controlled account deletion and recovery period, data minimization principles, and Standard Contractual Clauses for any cross-border data transfers. Audit logging captures all admin actions and sensitive data access for compliance reporting. The OWASP Top 10 is addressed through input validation, parameterized queries, CSRF protection, Content Security Policy headers, and rate limiting on all public endpoints.

Tech Stack

Next.js (React) with TypeScript for the frontend and SSR layer. Deno microservices (TypeScript) for AI orchestration, PDF generation, payments, and notifications. Prometheus for metrics collection across all services. Grafana for dashboards, alerting, and distributed tracing visualization. Stripe for payment processing. AWS S3 for PDF storage and delivery. OpenAI and Google Gemini APIs for AI generation. Cloudflare for edge security and performance. Vercel for frontend hosting. Custom admin panel with role-based access control.