Prestilien Pindoh
Brussels, Belgium · Remote
AI/ML and software engineer owning the path from model and agent architecture to cloud delivery, evaluation, and observability.
AWS certified
Cloud-native delivery
Founder & CTO
Product ownership
Build → deploy → observe
Production mindset
AI Engineering
I design reliable workflows around models: explicit state, grounded context, human control, and a path to evaluation.
A multi-agent career platform that turns a job search into a controlled workflow instead of a sequence of disconnected prompts.
System flow
An academic assistant that grounds answers in course material, shows its sources, and turns documents into active study workflows.
System flow
Selected work
ML pipelines, SaaS products, and client platforms—evidence that I can move between research, backend systems, and user-facing delivery.
Product engineering
A culinary product connecting people with authentic African recipes across web and mobile.
Founder-built product
Business software
Sales operations software spanning APIs, relational data, object storage, and automated delivery.
Full-stack ownership
Production ML
An anomaly model taken beyond the notebook into tracked experiments, containerized inference, and automated delivery.
Applied ML
An unsupervised learning pipeline that turns purchase behavior into interpretable customer segments.
Client delivery
A bilingual company website with localized content, project storytelling, and a direct quote flow.
Experience
I have worked across applied AI, founding teams, and full-stack delivery—so I can reason about the model and the product around it.
AI
Applied systems
0→1
Product building
E2E
Delivery ownership
Multitel
Explored reinforcement learning and explainability for intelligent systems, with reproducible experiments and production constraints in mind.
Afrik Delices
Owned product direction and engineering for a food-tech platform, from user experience to full-stack and mobile delivery.
Surfyn
Delivered web, mobile, and backend features that connected polished interfaces to dependable services.
VALIDE
Set technical direction and evolved an education product from early architecture to AI-assisted learning experiences.
Production engineering
A model is only one dependency. I think about the API, deployment, failure recovery, monitoring, and feedback loop around it.
System boundaries, failure modes, model behavior, and human checkpoints.
Testable workflows, typed APIs, retrieval pipelines, and evaluation loops.
Containers, CI/CD, infrastructure as code, and cloud-native delivery.
Metrics, traces, errors, latency, cost, and LLM output quality.
AI-native workflow
I use Cursor, Claude, and OpenAI across research, implementation, debugging, code review, and documentation—while tests, observability, and human review remain the quality gates.
Production toolkit
Credentials
Cloud engineering and agentic AI learning backed by public credentials.
I'm looking for AI/ML or software engineering work where strong systems thinking, product judgment, and production ownership matter.
