About Me

I am a freelance ML & AI engineer with a strong background in agentic systems, MLOps, distributed systems, and cloud-native architectures.

I build reliable AI applications and support their entire lifecycle - from early prototypes to large-scale production systems. My work involves both hands-on engineering and mentoring teams.

Active open-source contributor and co-founder of gradion.ai.

Professional Journey

I’ve helped companies like MerlinOne and Canto build and deploy large-scale AI search platforms that process and index millions of images and videos in real time, combining distributed inference (PyTorch, Triton, Kubernetes) with distributed vector databases (Milvus) for scalable, low-latency retrieval. These platforms are used by thousands of customers including The Associated Press - enabling them to use advanced visual image and video search simply by describing visual content. I also scaled computer vision models on AWS to process over 100M+ assets and conducted model fine-tuning to optimize search performance on domain-specific data.

At cyan Security Group, I designed a comprehensive MLOps platform (Databricks, MLFlow, AWS) that automated data pipelines, model deployment, and real-time monitoring —allowing rapid iteration from prototype to production. I also trained and integrated computer vision models into the cyan security suite, including child protection and content filtering products, to detect malicious or sensitive web content.

As a distributed systems engineer, I have developed a reactive microservices framework and built a distributed messaging infrastructure to enable high-throughput, event-driven communication—ensuring scalability and fault tolerance across mission-critical client services.

Skills & Expertise

• AI Engineering - Implementation of AI systems from RAG solutions to LLM-based agentic systems.
• Machine Learning – End-to-end model development, training, fine-tuning, and deployment.
• MLOps & Platform Engineering – Building robust pipelines that streamline data management, model development, deployment and monitoring.
• Distributed Systems & Architecture – Designing reliable, fault-tolerant infrastructures that efficiently scale with workload demands.

Technologies

Below are some of the technologies and tools I use in my day-to-day work:

🤖 AI & Machine Learning

PyTorch, Hugging Face (Transformers, Datasets, Accelerate, TGI), Scikit-learn, LangChain, Haystack, llama.cpp, Triton, Faiss, Milvus, Weaviate, Qdrant

🧠 Model APIs

Anthropic, OpenAI, Gemini

🛠️ Application Development & Data Engineering

FastAPI, Flask, Apache Kafka, Apache Spark, Delta Lake, Databricks

🌐 Infrastructure

Docker, Kubernetes, AWS, Azure, Google Cloud Platform, Terraform, GitHub Actions