Jean-Dominique Michaël Casanova, Founder of Drafted By: EdTech, FinTech and AI engineer
Founder, Drafted By
EdTech · FinTech · AI
I build and run AI products for two concrete jobs: planning lessons across 13+ national curricula and producing retirement audit reports for French advisory firms.
Retirement audit SaaS, France
- carriva.fr: RAG audit of French career records (RIS): parsing, anomaly detection, audit reports
Lesson planning AI, four markets
- preparemescours.fr: French curriculum
- draftmylesson.com: English-speaking curricula (UK, US, AU, CA)
- przygotujlekcje.pl: Polish curriculum
- creaclases.com: Spanish curricula (Spain + 9 LATAM)
Teacher planner
- agenda.preparemescours.fr: free planner for teachers (timetable sync, classes, tasks); PMC-powered lesson scheduling on top
Homework diary
- cahier.online: one permanent link a teacher shares with students, per-class homework and due dates for the year
Chess workspace
- chessdesk.app: all-in-one off-the-board workspace for coach and student, free Stockfish game review and analysis tools
1-on-1 tutoring
- jdteachai.com: 1-on-1 in AI tools, English & French, for professionals
Founder & coach, JD Chess
1-on-1 coaching · group masterclasses · digital tooling
I coach 1-on-1, run group masterclasses, and publish the supporting material. Structured training plus per-game analysis takes most students up to +500 Elo in 10 weeks.
Key Achievements
- Site & booking: FR + EN, lesson archive, testimonials. Built and shipped solo. (jdchess.com)
- Volume: 5,000+ individual sessions delivered worldwide. Verifiable Elo gains per student.
- Ranking: Top 3 chess coach in France on Superprof, 5-star average.
- Group program: Launched JD Chess Masterclass: structured group format, theory + live practice.
Head of European Automation
Support automation · ML · Project Management
I led automation for WhatsApp's European support operations: ML deployments, vendor delivery, and the internal support platform. The work cut manual ticket handling by ~50%.
Key Projects & Impact
- Automation: As Head of European Automation, cut manual ticket handling by ~50%.
- ML: Shipped ticket-closure ML models. Cut manual work and response time by 10%.
- Platform: Built WASP, the internal support portal: ticket management plus automation.
- Misinformation: Privacy-preserving analysis of misinformation patterns in French-speaking Africa.
- Interop: Worked with legal and engineering on cross-app messaging interoperability across the Meta family.
Business Network Engineer
B2B data exchange · EDI infrastructure
I supported enterprise clients on the GXS/OpenText Trading Grid, tracing B2B data, security, and transmission failures across global supply chains.
Technical Responsibilities
- Network migration: Zero-downtime migrations for critical trading partners under strict security compliance.
- EDI architecture: Trading partnerships and EDI-boxes across global supply chains.
- Forensic debugging: Granular data tracking across ANSI X12 and EDIFACT. Caught transmission failures before client impact.
- Protocols: AS2, SFTP, OFTP2 connectivity, gateway interoperability.
Technical Support Senior Advisor
Escalation management · mentoring · documentation
I owned advanced cases on Apple's technical escalation desk, mentored junior advisors, and wrote troubleshooting guides for the team.
Key Contributions
- Escalations: Owned advanced hardware and software cases for high-value clients.
- CSAT: Top-percentile customer satisfaction scores across tenure.
- Mentoring: Coached junior advisors on troubleshooting and customer-facing communication.
- Documentation: Authored troubleshooting guides used across the support team.
Early client-facing roles
Media · sales · service
At France TV, Europcar, and Orange, I learned client work on the floor: listen closely, explain clearly, and solve the problem in front of you.
Core Competencies Developed
- Client work: Sales and service across multiple verticals.
- Media: Communication and audience-facing operations at France TV.
- Communication: Active listening and clear writing, habits I still rely on.
Cinema, economics, and sciences
Cinema · economics · sciences
Cinema taught me narrative structure, economics taught me incentives, and sciences taught me to test a hypothesis. I still use all three when I build or debug a system.
Fields of Study
- Cinema: Visual storytelling, narrative structure, audience engagement.
- Economics: Complex systems, incentives, data analysis.
- Sciences: Hypothesis-driven thinking applied to troubleshooting and process design.
Work with me
Product, automation, and teaching projects
I can help with solo product delivery, RAG-backed software, support automation, and structured teaching. Tell me what you are building or fixing, and I will reply by email.
Software studio: draftedby.com · Retirement SaaS: carriva.fr · Private lessons: jdteachai.com · Chess coaching: jdchess.com · Workshop notes: jdmcasanova.com/blog
The Kernel
I turn unclear problems into testable parts, then trace each part to its cause.
- Problem Solving: Systematic diagnosis that traces a symptom to its cause.
- Process Automation: Workflows that remove repeated manual steps.
- Strategic Planning: Long-term direction broken into milestones you can verify.
The Interface
I translate technical decisions into language that clients, teachers, and operators can use.
- Cross-Cultural Communication: Adapting tone, examples, and stakes per audience.
- Escalation Management: Turning high-pressure conflicts into structured resolutions.
- Technical Writing: Docs that let the next person complete the work without guessing.
The Application
Chess coaching and support operations use the same loop: explain the task, practise it, review the result, and adjust.
- Structured Teaching: Breaking down complex skills into learnable steps.
- Performance Tracking: Metrics that guide improvement, not decoration.
- Mentorship: Building independence, not dependency.
System Output
The useful output is live, measured, and still running. These are the tools and results behind the work.
Tooling Stack
Key Metrics
- Products: 7 sites in production: a 4-market EdTech suite, retirement SaaS, AI tutoring, and a chess academy.
- Automation: ML-driven ticket routing reduced manual handling by ~50% at WhatsApp.
- Infrastructure: Self-hosted production stack on Proxmox: Docker, Cloudflare Tunnels, PostgreSQL, Meilisearch.
- AI/RAG: LangChain + OpenRouter for document analysis, lesson generation, and Carriva's retirement audit pipeline.
Field notes
Latest from the workshop
Build notes from production EdTech, retirement software, and AI tools: exact changes, failures, and operating trade-offs.
5 min readMigrating a Next.js portfolio off Vercel to Coolify
Cutover steps from Vercel to a homelab Coolify, what broke, and why operational coherence beat the symbolic cost savings.
vercelcoolifycloudflare-tunnelnextjssolo-founder
6 min readWhat I changed in next.config.mjs when leaving Vercel
Three diffs that earn their place: output standalone, edge runtime switch, explicit compression. Plus the security headers you stop getting for free.
nextjsverceldockercoolifysolo-founder
4 min readGitHub Deploy Keys Are Unique by Public Key, Not by Repo
I tried to reuse an SSH deploy key across two repos. GitHub returned a 422. The key is globally unique per user account, not per repo.
githubdevopssolo-foundergotcha
5 min readThe Coolify monorepo webhook that silently skipped 3 of 4 apps
A commit to shared CSS across four Next.js apps in the same monorepo only deployed to one. The other three reported 'finished' but served stale code.
coolifymonorepodeploygotchasolo-founder