Youssef Jaafar · Senior AI/ML Engineer
I make LLM & voice systems reliable enough to ship in regulated industries
I build production-grade chat and voice AI for regulated industries like healthcare, insurance, banking, and law, combining LLMs, STT/TTS pipelines, RAG, and agentic architectures with the evals, guardrails, and observability that keep them dependable in production.
AI/ML solutions shipped to production
Engineers led & mentored
STT accuracy gain (WER-benchmarked)
CSAT gain on agentic rollout (post-call survey)
Tools, Platforms & Languages I Build With
About Me
Building reliable AI systems that scale for enterprise environments
Core Focus Areas
Senior AI/ML Engineer specializing in enterprise Conversational AI, LLM systems, and AI-powered customer experience platforms.
I have extensive experience delivering solutions for regulated industries, with ownership of mission-critical AI systems supporting executive stakeholders in global, confidential environments.
My expertise spans the full AI development lifecycle, from data pipelines and model evaluation to production deployment and continuous optimization.
What Sets Me Apart
Engineering-first approach to LLM reliability
Focused on building systems that work consistently in production with measurable quality metrics.
Systems thinking: from data + eval to deployment
End-to-end ownership from data pipelines through model evaluation to production deployment.
Deep conversational AI + speech expertise
Specialized knowledge in voice bots, STT/TTS optimization, and multi-modal AI interactions.
Experience
A track record of delivering enterprise AI solutions at scale
Senior Software Development Engineer
Conversational AI & LLM Systems
Brillio (Bain Capital Company)
- Tech lead for 15+ engineers: coaching, code reviews, and documentation best practices
- Led 3+ production-ready agentic AI solutions in the last 6 months; improved CSAT by 50%+ (post-call survey)
- Ownership of agentic AI feature design and development
- Lead developer across 3 conversational AI projects using hybrid agents + assistants
- Partnered with IBM & GCP product teams to resolve platform issues and unblock critical AI capabilities
- Improved LLM reliability (hallucination reduction) via prompt engineering, evaluation strategies, and controlled reasoning patterns
Customer-driven US relocation; full citation in Recognition below.
Senior Software Development Engineer
Conversational AI & LLM Systems
Brillio
AI Software Engineer
Cedrus Digital (acquired by Brillio)
Featured Projects
Production-ready AI systems built for enterprise environments
Astra — AI Copilot for a Personal Life OS
May 2026 – Present
AI copilot that reasons over your own tracked data for grounded, personalized insights
AI Agent Evaluation with DeepEval
Jan 2026 – Present
LLM evaluation framework for IBM & GCP AI agents using DeepEval
Agentic AI Systems & LLM Applications
2024 – 2025
Multi-agent LLM systems for automated reasoning and tool execution
Knowledge Graph & RAG Platform for SEC Filings
2025
Natural-language querying over structured financial documents
Caristmatic AI - Car Price Prediction
May 2024 – August 2024
ML-powered vehicle price estimation based on historical data
E Healthcare Management System
January 2021 – May 2021
Full-stack online consultation platform for healthcare providers
Data Science for Climate Change - Wheat Yield Prediction
August 2021 – December 2021
ML/DL pipeline for agricultural yield forecasting
Text-Based Emotion Detection
August 2020 – December 2020
NLP system for sentiment analysis using neural networks
Wind Speed Prediction
February 2021 – May 2021
Time series forecasting for renewable energy applications
Open Source & Client Contributions
Active development across enterprise and open source projects
last year
How I Architect Conversational AI
A representative production pipeline: speech in, agentic reasoning with retrieval and guardrails, speech out, observable and continuously evaluated end to end.
Tap any block to explore the stage, its tech, and a key tradeoff.
Agent Orchestrator
Planning · RoutingRole
The brain: classifies intent, plans steps, and coordinates capabilities and handoffs.
Tech
Agent framework, intent routing, state/memory, tool calling
Tradeoff
Hybrid design: agents for open-ended reasoning, deterministic flows where compliance demands predictability.
Skills & Expertise
Comprehensive technical toolkit for building enterprise AI systems
Core Strengths
LLMs & Agentic AI
Building multi-agent systems with AutoGen, LangChain, and advanced prompt engineering
Conversational AI
Enterprise chat and voice solutions for regulated industries
Speech (STT/TTS)
Production pipelines with measurable accuracy improvements and scale
Production Deployment
Delivering reliable, scalable AI systems to production environments
System Reliability
Incident resolution and maintaining mission-critical AI infrastructure
Recognition & Awards
Formal recognition for consistent excellence, leadership, and impactful contributions
Brillio Star Award
Individual Excellence
Presented on: September 2025
Awarded by: Harish G (Current Manager)
Youssef has been a tremendous asset to the YAVA project. His consistent high performance has earned the appreciation of the customer, so much so that they requested his relocation to the US to continue contributing to the team.
He has played a key role in resolving critical issues in the Speech Model product and quickly upskilled himself to work on new platforms like Watsonx Orchestrate and LSM. His ability to adapt and deliver results is truly commendable.
Youssef has also gone above and beyond by leading and demonstrating key capabilities across Claims, YF, and FC to IBM, showcasing exceptional ownership and technical depth.
Great job, Youssef, your dedication and impact are truly valued!
Group Excellence Award
Team Excellence
This certificate of appreciation is awarded for truly demonstrating Excellence.
Presented: August 2024
By: Brillio Management Team
Group Excellence Award
Team Excellence
This certificate of appreciation is awarded for truly demonstrating Excellence.
Presented: February 2024
By: Brillio Management Team
What Colleagues Say
Recognition and appreciation from leadership
Since joining the YAVA program, I've been genuinely impressed by how Youssef manages so many competing priorities. He brings dedication, positivity, and a sense of grace that is rare & deeply appreciated. Youssef not only supports the team - he strengthens it, inspires it, and consistently contributes to our success in meaningful ways. Thank you Youssef!
Colman, Susan
We deployed yesterday with a successful E2E Checkout. Special effort as it required collaboration with multiple teams like MCOE, Five9 and coordinating overall with MWB Team. Also, heavy collab with IBM Team. Special call out to Youssef for chiming in for bug fixes and giving so much time.
Shreyank, Patel
Thank you so much for your outstanding support in the whole day working session today on the E2E integration of YAVA First and the Intent Agent for FAQ and Benefits Agents. Every one of you was instrumental in resolving the pending items.
Subramanian, Viji
THANK YOU Youssef, for your help with demoing Claims, YF and FC to the IBM team today :) I know discussing the FC development at 5pm on a Monday is not the most fun task so I appreciate your support!
Adwan, Meis
Thanks to all for 'tirelessly' working through the efix tasks today and for your great attitudes and support. Best team ever!!
James, Kathy
Thank you so much team explorers for awesome work that was delivered for August especially. Lots of challenges/roadblocks that came way but team did a great job in accomplishing this milestone for August release.
Easwaran, Shiva
Services
Enterprise-ready AI solutions tailored to your business needs
Conversational AI Solutions
Chatbots, voice bots, IVR replacement, AI customer service design and deployment.
LLM + Speech AI Systems
End-to-end pipelines using LLMs, STT, TTS, knowledge graphs, and RAG.
Agentic AI & Tool-Using Architectures
Agent design (roles, planning, orchestration) and safe tool integration for business workflows.
AI Quality Assurance
Hallucination evaluation, reasoning consistency checks, WER benchmarking, regression testing.
Regulated AI Deployments
Architecture and delivery for regulated industries with reliability, auditability, and operational readiness.
Continuous Optimization
Data-driven evaluation loops, monitoring, quality metrics, and iterative improvements.
How I Work
Education
Foundations in engineering and data science
B.E. Computer & Communications Engineering
Data Science Track
American University of Beirut • 2018 – 2022
Hobbies & Interests
What I enjoy outside of work
Reading & AI/ML Research
Keeping up with latest AI/ML technologies, research papers, and industry trends
Judo
Practicing the martial art of Judo for discipline, fitness, and mental strength
Cycling
Exploring new routes and staying active through regular cycling sessions
Mindfulness Meditation
Practicing mindfulness for mental clarity, focus, and overall well-being
Get in Touch
Let's build reliable AI systems that scale
Let's Connect
Whether you're looking to build enterprise AI solutions, need consultation on conversational AI systems, or want to discuss potential collaborations, I'd love to hear from you.