Senior Backend EngineerDubai, UAE
I’m Wilfred Chetat. I work mainly in Python on APIs, integrations, payments and data systems, and I build products like Intrvuer from interface to deployment.
Each project is labelled with how I was involved.
A voice interview coach that ties practice to the role you’re preparing for.
Candidates add a résumé and target job, answer questions out loud, and get feedback on each answer.
My contribution: design, Next.js interface, FastAPI backend, speech and LLM integrations, deployment.
Facebook Lead Ads to Adobe Campaign, from webhook to sync status.
I built the lead integration end to end: webhook processing, field mapping, sync status and connection validation, plus audience-management APIs and the React screens around them.
My role: Freelance Full-Stack Engineer · Upwork
| Lead form field | Campaign field | Status |
|---|---|---|
| full_name | [Campaign field] | Mapped |
| [Campaign field] | Mapped | |
| phone_number | Not mapped | Needs a field |
Leads sync once required fields are mapped.
Data ingestion for lease-insurance operations.
Source files had inconsistent periods, repeated transactions and running totals that couldn’t be trusted. I worked on ingestion rules, the relational model, APIs, duplicate prevention and client-scoped access.
Opportunity ingestion for African talent.
I worked on the opportunity-ingestion backend, turning source listings into consistent records through per-record validation and replay-safe upserts, with an outcome for every record: created, updated or rejected.
Smaller builds for learning and testing ideas.
A custom tokenizer and parser supporting nested arrays and objects, typed JSON values, structured errors and formatting, developed with cohort feedback.
{
"name": "Ada",
"tags": ["rust", "json"],
"active": true
}Object
├ name: String "Ada"
├ tags: Array
│ ├ String "rust"
│ └ String "json"
└ active: Bool trueFind a detection by describing clothing or belongings, using generated descriptions, text embeddings and Qdrant retrieval. I built the backend and search interface. Separate from production work.
Plain explanations of tools I use.
How embeddings and a vector database find visually similar images, with a FastAPI, Qdrant and InsightFace demo.
My work spans payment and wallet systems, data ingestion and video analytics. Most of it is in Python, and I also build the interfaces and integrations that turn a backend into a usable product.
Alongside employed roles, I take on freelance client work and contribute to EduClick’s opportunity-ingestion backend. I also built Intrvuer, my own product.
For an engineering role or a freelance project involving Python backends, integrations, data workflows or AI, send a short outline of what you need.
yekuwilfred@gmail.com