The Enterprise AI Architect Workshop (3 Days)
The Enterprise AI Architect Workshop
Enterprise AI is not a ChatGPT subscription. It is a complex architectural challenge requiring models that run locally, integrate deep business context, and orchestrate multiple agents to automate real-world processes.
To achieve this, you need people capable of building it. This workshop is driven by three critical requirements:
- Controllable Costs: Stop relying on "black box" API pricing. We teach self-hosting to move from volatile token fees to predictable, cheaper infrastructure costs.
- Data Privacy: For non-US-based enterprises, data sovereignty is non-negotiable. We architect systems where your data never leaves your perimeter.
- Absolute Security: Avoid the risk of third-party data leaks. By building a self-hosted, independent stack, you ensure your proprietary data is never exposed by another company's security failure.
3 DaysIntermediateRemote
Full Description
Agenda: Enterprise AI Architect (3 Days)
Basic Understanding of LLMs
- Understanding Pre-Training by simple examples
- Understanding Inference by simple examples
- Security & Privacy (Prompt Injection)
- Cost Structures
- Direct ChatML
Local LLMs
- Infrastructure: Hardware architecture, GPU/VRAM requirements, and Quantization
- LM Studio / vLLM
- WebLLM / transformer.js
Extending LLMs
- In-Context Learning
- RAG
- Fine-Tuning: BEIR
Agents
- Tool Calling
- Vendor-specific Development (Claude, GPT)
- Google Agent Development Kit
- MCP (Deferred Tool Calling)
- Agent Skills
- Agent Memory
- Harness Engineering: Security, Authentication, and Guardrails
- Evals
