Evaluation before rollout
I define quality checks up front (task accuracy, failure patterns, and acceptance thresholds) so AI features are shipped against evidence, not demos.
Software + AI engineering
Yassine El Ouni
Cloud-native platforms, APIs, and real-time systems shipped to production for international teams.
TypeScript, Node, NestJS, Angular, GCP, Firebase · pragmatic LLM integration and AI-assisted engineering where it earns ROI.
Platform & APIsAI & LLMsTechnical leadership
6+
Years delivering
Enterprise + startup
Contexts
Worldwide
Remote (EU, Americas, MENA, APAC)
Based in Tunisia (GMT+1). Remote-first for global teams — comfortable with overlap across Europe, the UK, MENA, North America (US & Canada), and Asia-Pacific (e.g. Australia) and similar zones.
About
Senior Software Engineer and AI consultant with 6+ years shipping cloud-native apps, scalable backends, real-time systems, and full-stack products. I optimize for maintainable architecture, measurable performance, and delivery that matches what stakeholders actually need.
Day-to-day AI engineering: Cursor-style assisted development, multi-model LLM workflows per task, and agent-style support for exploration and review—without lowering the production bar.
What I deliver
Systems and software engineering first; AI where it creates clear product value.
Build and scale
Production-grade backends, APIs, real-time systems, and cloud-native architectures on GCP and related stacks. Reliability, performance, and cost-aware operations for products that must hold up in the real world.
AI and intelligent features
LLM integration, intelligent workflows, and automation inside real products—not slide decks. Pragmatic patterns for prompts, evaluation, and guardrails, aligned with user trust and business constraints.
Lead and deliver
Technical leadership for squads and consulting engagements: architecture decisions, stakeholder alignment, and predictable delivery across enterprise and startup contexts.
AI in production standards
Practical rules I follow to ship AI features that are measurable, reliable, and safe under real usage.
Guardrails and fallback paths
Every AI flow has constraints (input/output checks, policy boundaries, and fallback behavior) to keep user experience stable when models drift or fail.
Observability and traceability
Prompt versions, model choices, and key outcomes are tracked to make regressions visible and speed up incident debugging.
Cost and latency discipline
I tune model usage, caching, and orchestration patterns for predictable response time and cloud cost under real traffic.
Human review where risk is high
For sensitive decisions, AI assists but does not finalize alone. Human-in-the-loop review is explicit in the workflow.
Incremental release strategy
I ship in stages (pilot, monitor, expand) with rollback plans, so AI capabilities improve safely in production.
Projects
Selected projects
Representative technical work from client and product initiatives; details summarized or anonymized where appropriate (same selected projects as on my CV).
AI B2B Lead Discovery Agent Platform
AIPlatform
- Built an AI-powered lead generation platform that generates ICPs, discovers leads, enriches contact data, and scores prospects.
- Designed an end-to-end workflow covering targeting, discovery, enrichment, and prioritization.
- Automated a large part of outbound lead research and improved operational efficiency.
LLM workflowsAgentsData enrichmentB2B automation
Autonomous AI Video Factory
AIPlatform
- Built a fully automated AI content production system for YouTube Shorts, TikTok, and Instagram.
- Designed a multi-agent workflow for ideation, script generation, visual generation, voiceover, and optimization.
- Automated the pipeline from content idea to distribution-ready output.
Multi-agentContent pipelineMedia generationSocial formats
Enterprise Foot Traffic Analytics Platform
AICloudReal-timePlatform
- Built an AI-powered analytics solution using computer vision for crowd flow and demographic analysis.
- Designed scalable backend services and distributed processing workflows.
- Delivered real-time analytics with a privacy-conscious architecture for enterprise use cases.
Computer visionDistributed processingReal-time analyticsPrivacy-aware design
Client engagements
Named roles and contexts—AI-led delivery first, plus platform and enterprise depth. Tags match the themes in Selected work & outcomes.
More detail on request.
Business-critical web & mobile products
ZUPdeCO
Senior consultant · AI & product technical strategy
2025 – present · France, remote
AILeadershipPlatform
Contract engagement leading design and delivery of products where speed, architecture, and AI-enabled workflows must coexist.
- Technical and architectural ownership with a strong product and ROI lens.
- AI-assisted delivery (Cursor-style workflows, multi-model usage) for faster iteration without lowering production standards.
- Shipped intelligent workflows, automation, and data processing alongside core application features.
TypeScriptNodeCloudAI toolingProduct delivery
Cloud-native, AI-enabled applications
Tested4you Business
Senior consultant · AI & cloud-native apps
2024 – present · France, remote
AICloudReal-timePlatform
Building and evolving applications where real-time behavior and cloud-native patterns matter under changing requirements.
- Designed and delivered cloud-native web and mobile experiences with Firebase and Firestore.
- Advised on architecture, scalability, and quality in a high-churn delivery environment.
- Focused on production readiness and performance under real user load.
FirebaseFirestoreTypeScriptReal-timeGCP
Retail cloud platform engineering
Carrefour Links (via Devoteam)
Senior consultant · full-stack & cloud
2023 · France, remote
CloudPlatform
Production features and platform work on a cloud-based retail stack with tight reliability expectations.
- Full-stack delivery on cloud-hosted services with GCP operations ownership.
- Improved data access and maintainability using Prisma ORM; pushed stability and cost awareness.
- Contributed to continuous improvement of delivery quality and production reliability.
GCPPrismaTypeScriptFull-stackPostgreSQL
Enterprise backend & technical assessment
PwC UK (via Play Studios LTD)
Senior consultant · backend & solutions
2023 · UK, remote
PlatformLeadership
High-performance backend work and structured technical communication for enterprise stakeholders.
- Delivered backend services with emphasis on reliability and clarity.
- Ran technical assessments and feasibility work to de-risk solution design.
- Documentation and alignment practices aimed at long-term maintainability.
NodeEnterprise APIsDocumentationStakeholder delivery
Selected work & outcomes
Same person—filters only change emphasis. Tags highlight themes across one track record.
Featured — AI & modern delivery
AI-Assisted Engineering and Intelligent Workflows
AIPlatformLeadership
In recent engagements, combined rigorous software delivery with modern AI practice: AI-assisted development (including Cursor-style workflows), multi-model LLM usage matched to each task, and agent-style automation for exploration, review, and documentation—without compromising architecture or production standards. Built and integrated AI-driven capabilities including automation, intelligent workflows, and data processing alongside classic full-stack and cloud work.
Full-Stack Application Development
PlatformReal-time
Led the development of a comprehensive full-stack application using Angular and Node.js with NestJS, implementing real-time communication via Socket.IO. This project improved user engagement by 40% and reduced data latency by 60%.
Cloud-Based Solutions Architecture
CloudPlatform
Designed and implemented scalable cloud-based solutions using Google Cloud Platform (GCP), resulting in a 30% reduction in infrastructure costs and a 50% improvement in application performance.
Microservices and DevOps Implementation
PlatformDevOps
Spearheaded the transition to a microservices architecture using Docker and Kubernetes, improving deployment frequency by 200% and reducing mean time to recovery by 60%.
Cross-Functional Team Leadership
LeadershipPlatform
Led a team of 7 developers in an agile environment, successfully delivering 5 major projects on time and within budget. Implemented code review processes that improved code quality and reduced bugs by 40%.
Insights
Short write-ups on delivery, cloud, real-time systems, and AI engineering practice.
Organizations
Logos for recognition; scope and role varied by engagement.
Skills & tools
Core stack — what I reach for most weeks. Details by area below.
TypeScriptNode.jsAngularNestJSGCPPostgreSQLDockerFirebase
Browse tools by area (compact view):
AI and modern engineering practice
How I ship today: assisted development, model choice per task, and disciplined integration of LLMs into real products.
Cursor & AI-assisted IDE workflows Multi-model LLM usage (task-matched) LLM APIs & product integration RAG & agent-style orchestration Prompts, evals, and guardrails Python for AI & data glue
Get in touch
Open to senior engineering, technical lead, and consulting conversations. Send a short brief (role, stack, timeline) and I will reply with fit and availability.
Site URL for signatures:yassine-elouni-profile.web.app