Engineering Reliable Agents with Ragas
Moving beyond the "Vibe Check" by instrumenting LangChain agents with the RAG Triad metrics.
Jan 24, 202610 min read
AgentsTestingRagasLangChain
Senior AI Engineer & Technical Lead
Architecting
Intelligence.
Specializing in moving LLMs from prototype to production. I bridge the gap between research science and scalable product engineering.
Scroll
Capabilities
The Mind Map
AI Engineering
Designing and deploying production-grade AI systems. From model architecture to inference optimization.
LLMs
ML Pipelines
RAG Systems
LLM Ops
End-to-end LLM lifecycle management. Fine-tuning, evaluation, and deployment at scale.
Fine-tuning
Evaluation
Monitoring
Product Strategy
Bridging technical capabilities with market needs. Building AI products that users love.
Roadmapping
User Research
GTM
System Design
Architecting scalable, resilient systems that handle millions of requests with grace.
Distributed Systems
APIs
Cloud
Team Leadership
Growing high-performance engineering teams. Mentoring, hiring, and establishing engineering culture.
Mentorship
Hiring
Culture
Portfolio
Selected Works
Biometric Privacy & Vision
Researching neural network robustness against occlusions. Published comparative studies on "Masked Face Recognition" performance during the pandemic.
PyTorchComputer VisionResearch
Autonomous Edge Robotics
Developed a Reinforcement Learning stack for self-driving vehicles on NVIDIA Jetson Nano hardware. Deployed across 6 European countries.
Reinforcement LearningEdge AIIoT
Enterprise AI Architecture
Designing data governance frameworks and predictive maintenance use-cases for international automotive and pharma clients.
System DesignArchitectureStrategy
Writing
Thoughts & Essays
Data Mesh: The Hidden Engine of AI
Why your AI Agents are failing: They are trying to drink from a swamp. The model is not the bottleneck, the data architecture is.
Jan 10, 202615 min read
Data ArchitectureAI Agents
Prompt Engineering That Actually Works
Stop telling the model to 'Act as Steve Jobs.' The engineering reality of reliable prompting.
Dec 5, 20258 min read
AI EngineeringPrompts
Principles
Engineering Principles
The best AI systems are invisible. They do not demand attention; they just work.
My focus is building the boring, reliable infrastructure that makes this possible. I value simple architecture over clever code. Every system I build is designed to be read by humans, maintained by teams, and trusted by users.
I operate where research meets production. This is where complexity usually kills projects, and where simplicity creates the most value.