██████╗ ███████╗ ██████╗ ██████╗ ██████╗ ███████╗
██╔════╝ ██╔════╝ ██╔═══██╗ ██╔══██╗ ██╔════╝ ██╔════╝
██║ ███╗ █████╗ ██║ ██║ ██████╔╝ ██║ ███╗ █████╗
██║ ██║ ██╔══╝ ██║ ██║ ██╔══██╗ ██║ ██║ ██╔══╝
╚██████╔╝ ███████╗ ╚██████╔╝ ██║ ██║ ╚██████╔╝ ███████╗
╚═════╝ ╚══════╝ ╚═════╝ ╚═╝ ╚═╝ ╚═════╝ ╚══════╝
███╗ ██╗ █████╗ ██╗ ███╗ ███╗ ███████╗ ██╗ ██╗
████╗ ██║ ██╔══██╗ ██║ ████╗ ████║ ██╔════╝ ██║ ██║
██╔██╗ ██║ ███████║ ██║ ██╔████╔██║ █████╗ ███████║
██║╚██╗██║ ██╔══██║ ██║ ██║╚██╔╝██║ ██╔══╝ ██╔══██║
██║ ╚████║ ██║ ██║ ██║ ██║ ╚═╝ ██║ ███████╗ ██║ ██║
╚═╝ ╚═══╝ ╚═╝ ╚═╝ ╚═╝ ╚═╝ ╚═╝ ╚══════╝ ╚═╝ ╚═╝
🚀 Quick Start
Engineering manager and ML product engineer with 6+ years building AI systems at scale. At SoundCloud, leading the Recommendations team and connecting product, engineering and data — while staying hands-on with ML, agentic systems and AI infrastructure serving 200M+ users.
🛠 Tech Stack
languages:
primary: [Python, Go, Scala, SQL]
secondary: [Lua, Bash, TypeScript]
frameworks:
web: [FastAPI, Django, React]
ml: [TensorFlow, PyTorch, HuggingFace]
infrastructure:
orchestration: [Kubernetes, Docker]
workflow: [Airflow]
iac: [Terraform]
ci_cd: [GitHub Actions, Jenkins]
databases:
sql: [PostgreSQL, BigQuery]
nosql: [DynamoDB, Redis, BigTable]
ai_ml:
platforms: [MLflow, Kubeflow, vLLM, TGI, Gemini]
models: [LLMs, CNNs, Transformers]
techniques: [RAG_Systems, Function_Calling, Tool_Use, Quantization, Semantic_Search]
agentic: [MCP, ADK, A2A]
evaluation: [Offline_Evals, Online_Evals, LLM_as_Judge, Metrics_Design]
platform_dx:
apis: [REST, GraphQL, gRPC]
observability: [Prometheus, Grafana, Custom_Metrics]
experimentation: [A/B_Testing, Feature_Flags, Metrics_Analysis]
💼 Experience
- Owning all home page products at SoundCloud, acting as the link between PMs and engineers across iOS, Android, web and data
- Leading a team of 5 — hired 3 of them, building a culture of ownership, fast iteration and high technical standards
- Built a company-wide translation service with Gemini, ADK and a unified caching layer — platform-agnostic, scaling from 13 to 21 languages
- Running agentic CI workflows where Cursor Agents and Claude Code pick up Linear tickets, open PRs and move them to review automatically — 75% faster resolution
Acting as de-facto engineering manager from early 2025 — leading the Recommendations team, aligning engineers and PMs, and owning end-to-end delivery before the role was formalized in January 2026.
- Led end-to-end recommendations product delivery — from algorithm work to cross-team integrations and UX improvements for 200M+ users
- Built data infrastructure pipelines processing trillion-row datasets with BigQuery optimizations, reducing costs and improving recommendation performance
- Implemented LLM-powered recommendation systems with function calling, tool use, and persistent context for adaptive personalization
- Built production RAG systems with hybrid search over 100M+ tracks using semantic embeddings and metadata fusion
- Spearheaded internal MCP implementations for tool orchestration across SoundCloud's infrastructure
- Pioneered metrics-driven development, shaping technical strategy and architectural decisions
- Built dynamic ranking systems driving 20% improvement in CTR and listening time across recommendations
- Built and maintained 3 Airflow DAGs covering data ingestion, feature pipelines and model training workflows
- Managed ML infrastructure on Kubernetes — model serving, pipeline orchestration and infrastructure operations
- Led privacy infrastructure implementation, cloud migration and infrastructure modernization
Built production-ready music analysis systems for audio segmentation and content detection.
- Music Segmentation: Deep learning models using TensorFlow for chorus detection and structure analysis
- NLP Integration: HuggingFace T5 models for audio transcription and vocal detection
- Scale: Processing pipeline handling 10k+ audio files daily with 95% accuracy
- Infrastructure: End-to-end ML pipeline from audio ingestion to insights delivery
Research and development of deep learning models for AI music composition and audio processing using TensorFlow, PyTorch, and C++.
- Audio compression using VAE, Autoencoder, and spectral analysis techniques
- Auto-dynamics processing with real-time amplitude/frequency modulation
- Transformer-based audio generation models for music composition
- Virtual instrument creation pipeline from sound design to sample generation
- Model optimization achieving 60% processing time reduction while maintaining quality
Computer vision and audio processing solutions for multi-modal AI applications.
- Computer Vision: YOLOv4 implementation for object detection and tracking
- Audio Processing: Custom sound detection models and audio-visual synchronization
- MLOps: Pipeline infrastructure and annotation workflows for ML model training
🎓 Certifications
- Lean Product Management - Itamar Gilad (September 2024)
Deep Learning & AI:
- AI Product Management - Miqdad Jaffer, Maven (2025)
- Deep Learning Specialization - DeepLearning.AI (2020)
- Neural Networks and Deep Learning
- Improving Deep Neural Networks: Hyperparameter tuning, Regularization and Optimization
- Structuring Machine Learning Projects
- Convolutional Neural Networks
- Sequence Models
- Generative Deep Learning with TensorFlow - DeepLearning.AI (2021)
Cloud & Infrastructure:
- Google Cloud Platform Big Data and Machine Learning Fundamentals - Google (2021)
- Google IT Automation Professional Certificate - Google (2021)
- AWS Fundamental Course - AWS (2020)
Specializations: Audio Signal Processing, Music Information Retrieval
🏛️ Education
MSc, AI in Sound and Music | Universitat Pompeu Fabra
⏱️ Sep. 19 – Sep. 20 | 📍 Barcelona
Specialized Master's program at the intersection of artificial intelligence and music technology, covering Machine Learning, Deep Learning, Digital Signal Processing, Music Information Retrieval, and Computational Music Creativity for algorithmic composition.
Civil Engineering | University of Balamand
⏱️ Nov. 11 – Jun. 15 | 📍 Beirut
Bachelor's degree in Civil Engineering with specialized focus on Acoustics, covering structural design principles, materials science, and sound engineering applications in architectural and construction projects.
⚙️ Configuration
Language Support
{
"native": ["English", "French", "Arabic"],
"conversational": ["German", "Spanish"],
"programming": ["Python", "Go", "Scala", "SQL", "Lua"]
}
📄 License
This professional profile is licensed under the "Let's Build Amazing Things Together" license.
