Training Course: Building Agentic AI Infrastructure
- Understand the key components of an Agentic AI architecture
- Design AI systems using LLMs, vector databases, and orchestration tools
- Deploy scalable knowledge-driven workflows using agents
- Explore integration patterns for enterprise data systems
-
Module 1: What is Agentic AI?
- Definition and evolution
- Agent vs. traditional AI workflows
- Use cases and emerging patterns
-
Module 2: Infrastructure Foundations
- Architecture overview: Frontend, API, backend, orchestration
- Core components: LLMs, vector DBs, Neo4j/graph DBs, APIs
- Tooling: LangChain, LlamaIndex, Streamlit, FastAPI, n8n
-
Module 3: Agentic Workflow Design
- LLM prompt management and chaining
- Autonomous vs. human-in-the-loop agents
- Memory and context management
-
Module 4: Data Layer Design
- Choosing the right vector DB (e.g., Weaviate, FAISS, Qdrant)
- Enterprise knowledge graphs (e.g., Neo4j)
- RAG (Retrieval Augmented Generation) best practices
-
Module 5: Governance and Guardrails
- Ethical AI, PDPA/GDPR, and internal protection
- Monitoring, feedback, and usage control
- Fail-safe design and observable pipelines
-
Module 6: Case Study + Hands-on
- Deploying a sample Agentic AI app using Streamlit + LangChain
- Connecting it with internal APIs and PDF data sources
- Course slides (PDF)
- Agentic AI infrastructure blueprint
- GitHub starter repo with components
- Use case canvas + system checklist
- ✅ 3-hour Essentials Workshop
- ✅ 1 Trainer / Up to 20 Participants
- ✅ First Three Module
- ✅ Digital Toolkit PDF
- ✅ Full-Day Interactive Workshop
- ✅ All 6 Modules (Comm + Collab + Social)
- ✅ 2 Trainers / Up to 30 Participants
Duration: 1 Day (6 hours)
Audience: CTOs, Data Architects, AI Engineers, Solution Consultants
Format: Live session + Hands-on + Templates
Level: Intermediate–Advanced
🎯 Course Objectives
📚 Course Modules
📦 What You Get
💼 Training Packages & Pricing
Starter
฿18,000
Professional
฿35,000
📞 Interested in Hosting This Course?
Contact us to run this training in-house or as a public session.