Workshops · Consulting · Agentic Engineering

Build Real AI Skills.
Ship Real Products.

Hands-on agentic engineering for developers and teams: build production AI agents, wire them into real products with agentic UIs — A2UI and CopilotKit — and run them on infrastructure you control. Taught by a Google Developer Expert and senior architects.

AI Agents & MCP
Agentic UI · A2UI · CopilotKit
Agentic Coding & Harnesses
Production RAG
Local & Sovereign LLMs

Browse Workshops Get in Touch

The stack we teach & ship with

A2UI
CopilotKit
MCP
Claude
OpenAI
pi
React
Angular
Ollama
Hugging Face
Meta Llama
Google Gemma
Mistral AI

Agentic UI — Beyond the Chatbox

The next generation of AI products doesn't answer in paragraphs. It renders interfaces. We teach and build exactly that.

Generative UI, not just chat

Agents that stream real interface — forms, tables, charts, approval cards — rendered live in your product with the A2UI protocol instead of walls of text.

CopilotKit in your React or Angular stack

Embed copilots into existing Next.js, React, and Angular apps: shared state between agent and UI, frontend tool calls, and context-aware actions. We maintain CopilotKit for Angular.

Human-in-the-loop by design

Interrupt, approve, and steer long-running agents from the UI. We build the checkpoints, guardrails, and audit trails that make agents production-safe.

Protocols, frameworks & harnesses we build with

A2UI Protocol

CopilotKitMCPMCP AppsReactAngularClaude Agent SDKpi

How can my agent render real UI instead of walls of text?

That's exactly what A2UI is for — we've written about it:

UI streamed by agent

Let the agent stream declarative UI and render it natively in your app — Angular or React. We run this in production, even on a local 12B model.

Sources · Soverius Engineering Blog

Read the engineering blog

Sovereign AI — Local LLMs on Your Infrastructure

The 'Soverius' in our name is no accident: we help teams own their AI stack — private, compliant, and cost-predictable.

Your data never leaves the building

Run open-weight models on your own hardware or private cloud. No prompts, documents, or customer data sent to third-party APIs — GDPR and works-council friendly by design.

Production-grade local inference

From a single Ollama instance to vLLM serving with continuous batching, quantization (GGUF, AWQ), and GPU sizing — we cover the real ops: throughput, latency, and cost per token.

Open models that hold their own

Llama, Gemma, Mistral and friends — we show you when an open 8B beats a frontier API on cost, and how to fine-tune with LoRA on your own domain data using PyTorch and Hugging Face.

Hybrid routing done right

Sensitive workloads stay local, hard reasoning goes to frontier models. We build routing layers that pick the right model per request — with fallbacks, evals, and observability.

Ollama
vLLM
Meta Llama
Google Gemma
Mistral AI
Hugging Face
PyTorch

From Problems to Outcomes

We address the real challenges teams face when adopting AI — and deliver measurable engineering skills.

The Problems We Solve

AI hype, no production skills

Your team has watched the tutorials and read the blog posts — but nobody knows how to actually integrate Claude or GPT-4o into a TypeScript codebase that ships.

Engineers blocked by the 'how'

Developers understand the promise of LLMs and agents but lack structured training on RAG architecture, tool use protocols, cost management, and production deployment.

Gap between demos and real systems

Most AI courses stop at calling an API. They don't cover error handling, observability, vector databases, fine-tuning trade-offs, or the TypeScript patterns that make LLM features maintainable.

The Outcomes You Get

Working TypeScript + PyTorch code

Build real RAG pipelines, AI agents, and structured output parsers. Use the Claude API, OpenAI API, LangChain.js, and pgvector — all in exercises you take home.

Production-ready architecture

Learn deployment patterns, cost optimization (Claude Haiku vs Sonnet vs GPT-4o-mini), rate limiting, error recovery, and observability — the things that matter after the demo.

Capstone projects you can adapt

Each workshop ends with a fully deployed project: a RAG-powered Q&A system or a multi-tool AI agent — ready to adapt to your own product and use cases.

Our Flagship Workshops

Hands-on intensives designed for developers and teams ready to build production AI systems.

AI for Developers (1 Day Workshop)

Rainer Hahnekamp, Murat Sari

The Enterprise AI Architect Workshop (3 Days)

999 €

Murat Sari, Rainer Hahnekamp

Browse Workshops

Why Teams Trust Soverius AI

Practical AI training by practitioners who've shipped it.

Hands-on TypeScript Labs

Every exercise uses real TypeScript codebases — the same stack you use at work. No toy demos.

Production-First Approach

We teach patterns that ship: error handling, observability, cost management, and deployment strategies.

Industry-Recognized Trainers

Led by a Google Developer Expert and experienced enterprise architects with real production deployments.

Post-Workshop Support

Questions after the workshop? We're here. Get follow-up support as you apply what you've learned.

TypeScript + PyTorch

dual-stack workshops

Claude & OpenAI

hands-on API integration

2-Day Intensive

from foundations to production

Meet Your Trainers

Led by practitioners who've shipped production AI systems, Angular architectures, and enterprise platforms.

Murat Sari

Serial Entrepreneur, Software Architect & AI Engineer

Murat Sari is a serial entrepreneur and software architect who has founded several startups and shipped products across the automotive, gaming, and live-stage visualization industries. He studied computer science at TU Wien and has spent over a decade designing large-scale enterprise systems — from...

Founded several startups spanning automotive, gaming, and live-stage visualization — created patents in real-time rendering
Ogre3D core team member (2012–2017) — contributed the HLMS physically-based shading system
TypeScript, Angular, .NET Core, Node.js/NestJS, Python/PyTorch — full-stack production expertise
Years of experience as enterprise Angular trainer — lecturer at FH Salzburg & FH Kapfenberg, architecture workshops
Lead Software Architect for real-time analytics in the oil industry

Rainer Hahnekamp

Google Developer Expert for Angular & AI Trainer

Rainer Hahnekamp is co-founder and AI Engineer at Soverius AI, and a Google Developer Expert (GDE), with fifteen years building mission-critical enterprise systems. He specializes in local LLMs and everything it takes to run them in production — retrieval, harness and loop engineering, agents, and...

Google Developer Expert (GDE)
Conference speaker: Devoxx BE, AI-Poland, AI-India, ng-conf, Basta, JAX, WeAreDevelopers
https://www.youtube.com/@rainerhahnekamp
15+ years at Mars, Inc. — team lead for production Angular & Spring systems

Frequently Asked Questions

Answers to common questions about our workshops.

How is pricing structured?

We offer per-seat pricing for individual participants and volume discounts for teams. Pricing is quoted per workshop based on group size and delivery format (in-person vs. remote). Contact us for a tailored quote.

What do I need to know before attending?

Participants should have programming experience (Python preferred) and familiarity with basic software development concepts. No prior ML or AI experience is required—we cover foundations at the start of each workshop.

What should I bring to the workshop?

A laptop with a code editor, Python 3.10+, and the ability to install dependencies (pip, virtual environments). For remote workshops, a stable internet connection and a quiet workspace are recommended.

Are workshops available remotely?

Yes. Both flagship workshops can be delivered in-person or remotely. Remote sessions use collaborative tools and shared coding environments to maintain the hands-on experience.

Do you offer group or team discounts?

Yes. Teams of 5+ participants receive volume pricing. We also offer customized private workshops for organizations with specific learning objectives.

Ready to build real AI skills?

Join developers and teams who've transformed their AI capabilities with our hands-on workshops.