AI Architecture & Discovery
From use-case framing and C4 system design to delivery planning, integration constraints, and technical strategy.
Available worldwide
I design and deliver production AI systems.
LLM applications, agentic workflows, RAG platforms, multimodal systems, and custom computer vision solutions — from discovery and architecture to deployment, integration, and handover.
Projects
15+
AI systems delivered
Focus
LLM + CV
Agentic and multimodal
Work mode
Global
Consulting, contract, full-time
LangGraph
LangChain
vLLM
Ollama
Qdrant
Weaviate
Neo4j
Kubernetes
Computer Vision
Personal AI Practice
Architecture, delivery, and hands-on implementation for enterprise AI and industrial workflows.
AI Solutions Architect
Architecture-first AI delivery
Building production AI systems for real workflows, private environments, and long-term maintainability.
Services
From architecture to production delivery
I help teams turn AI capabilities into working systems aligned with business goals, infrastructure realities, and long-term maintainability.
LLM & Agentic Systems
RAG platforms, tool-using agents, multi-agent workflows, evaluation loops, and production-oriented orchestration.
Serving & AI Platforms
Self-hosted inference, open-source model serving, vector databases, Kubernetes-based deployment, and controllable AI infrastructure.
Computer Vision & Multimodal AI
Detection, segmentation, classification, and multimodal pipelines connected to operational reporting and decision support.
Selected work
Private delivery and public AI systems work
A mix of enterprise-style delivery and public technical projects that reflect architecture, implementation, and production thinking.
Private case · Oil & Gas
Core Analysis & Geological Reporting
Computer vision system for geological core analysis with detection, segmentation, and classification to identify potentially oil-bearing layers, plus an AI agent for structured geological reporting.
Computer Vision
Multimodal Workflow
AI Reporting
Enterprise Delivery
Private case · GovTech
Voice Assistant for Public Services
A multi-agent voice-enabled assistant for a public services platform, helping citizens navigate services, submit meter readings, book appointments, and complete workflows through conversation.
Multi-Agent
Voice UX
Workflow Automation
Public Services
Public project · RAG Platform
OpenRAG
A self-hosted RAG platform with six retrieval strategies, full pipeline tracing, A/B testing, and interactive knowledge graph exploration.
RAG
Self-Hosted
Qdrant
Neo4j
Open project
Public project · LLM Lifecycle
PulsarAI
A unified platform that trains, deploys, evaluates, and improves LLMs in one closed loop with visual DAG pipelines and agent orchestration.
SFT/DPO/GRPO
Agents
Evaluation
Serving
Open project
Process
How I work
Strategy, systems thinking, and hands-on execution across the full AI delivery lifecycle.
01
Discovery
Business goals, user workflows, deployment constraints, and success criteria.
02
Architecture
Practical system design with clear trade-offs, integration paths, and delivery planning.
03
Delivery
Fast movement from concept to MVP or pilot while keeping architectural clarity.
04
Productionization
Reliability, observability, deployment model, and long-term maintainability.
05
Handover
Documentation, knowledge transfer, and solution readiness for internal teams.
Stack
Technology and delivery toolkit
Production-oriented tools and systems I use across architecture, serving, retrieval, and integration.
LLM / Agents
LangGraph
LangChain
CrewAI
OpenAI SDK
Anthropic SDK
DSPy
Instructor
LiteLLM
Function Calling
Computer Vision
PyTorch
Ultralytics YOLO
SAM 2
OpenCV
Supervision
ONNX Runtime
TensorRT
Roboflow
Serving / Inference
vLLM
Ollama
TGI
Triton Inference Server
SGLang
Retrieval / Knowledge
Qdrant
Weaviate
pgvector
Neo4j
Redis
Backend / Infra
Python
FastAPI
Kafka
Docker
Kubernetes
Helm
Architecture / Delivery
C4 Model
UML
Solution Design
Technical Documentation
About
Architecture-first AI delivery for real systems
I work at the intersection of AI architecture, delivery, and hands-on implementation. My focus is turning AI capabilities into practical production systems: LLM platforms, agentic workflows, RAG systems, multimodal applications, and custom computer vision solutions.
I am especially interested in private and controllable AI infrastructure, open-source LLM ecosystems, and systems designed around real workflow constraints rather than demos.
Consulting
Contract Delivery
Technical Leadership
Full-time Opportunities
Contact
Let's build something useful
Available for consulting, architecture work, contract delivery, technical leadership, and full-time opportunities.