ML Systems Engineer and Researcher focused on adaptive inference infrastructure, reinforcement-learning-driven optimization, distributed AI systems, federated learning, and large-scale experimentation.

I build production-oriented AI infrastructure for:

• adaptive inference serving
• reinforcement-learning-driven routing
• distributed systems orchestration
• online optimization
• observability-integrated ML systems
• scalable experimentation platforms
• trustworthy and heterogeneous federated learning

Distributed adaptive inference infrastructure for reinforcement-learning-driven LLM routing, online optimization, and production-scale serving experimentation.

• FastAPI serving infrastructure
• Redis-backed distributed workers
• Kubernetes deployment manifests
• GPU-aware scheduling
• Prometheus/OpenTelemetry observability
• streaming inference APIs
• adaptive traffic shaping
• contextual bandits
• Thompson Sampling
• policy-gradient routing
• Pareto tradeoff analysis
• oracle-regret evaluation

Repository:
https://github.com/vicobarafor/EvalRouteOps

Technical Report:
docs/EvalRouteOps_Technical_Report.pdf

Adaptive federated learning infrastructure for personalized optimization under heterogeneous client environments.

• adaptive aggregation
• client personalization
• heterogeneous federated learning
• drift-aware optimization
• scalable federated experimentation
• distributed training infrastructure

Repository:
https://github.com/vicobarafor/FedAdaptOps

Research infrastructure for studying geometry dynamics and instability in federated LoRA systems under heterogeneous client distributions.

Repository:
https://github.com/vicobarafor/federated-lora-geometry

Drift-aware adaptive aggregation for federated learning under non-IID client environments.

Repository:
https://github.com/vicobarafor/robust-federated-learning-noniid

Research experiments investigating personalization depth and client adaptation behavior in federated learning systems.

Repository:
https://github.com/vicobarafor/federated-personalization-depth

• reinforcement learning systems
• distributed AI infrastructure
• adaptive inference serving
• LLM routing systems
• online optimization
• Kubernetes orchestration
• observability systems
• ML systems engineering
• scalable experimentation infrastructure
• federated learning
• trustworthy AI systems

My work emphasizes:

• production-oriented infrastructure
• reproducible experimentation
• systems-level optimization
• observability-first design
• scalable distributed execution
• benchmark-driven evaluation
• adaptive online learning systems

https://victorobarafor.com

Currently exploring:

• reinforcement-learning-driven serving systems
• adaptive traffic allocation
• distributed inference orchestration
• infrastructure-aware optimization
• scalable evaluation systems
• trustworthy federated learning
• online adaptive AI systems