Understand, design, and deploy autonomous AI agents aligned with business goals and data ecosystems
This course allows participants to explore the different categories of AI agents, such as virtual assistants, conversational agents, and autonomous agents. It supports them in defining a clear roadmap for deploying AI agents adapted to the company’s needs.
Is it for you ?
Executives, project managers, consultants, and innovation leaders.
Prerequisites
Foundations in artificial intelligence and knowledge of business needs.
What You'll Walk Away With
- ✓ Identify and prioritize high-impact AI agent use cases using structured analysis
- ✓ Design functional architectures integrating memory, tools, and data access (RAG)
- ✓ Prototype agents that process requests and trigger automated actions
- ✓ Select the right frameworks and technologies based on autonomy and integration needs
- ✓ Deploy and monitor AI agents using KPIs, tracking tools, and continuous improvement practices
Training content
1 Introduction to AI Agents
- Definition of an agent and differences from a traditional LLM.
- Key concepts: autonomy, objectives, memory, ability to act.
- Overview of market technologies.
2 Short workshop
- Identify an agent in one’s professional environment (existing or potential).
3 Types of Agents
- In‑depth analysis by category:
- Conversational agents: FAQ, support, onboarding.
- Autonomous agents: report generation, complex automation.
- Multi‑tool agents: able to use APIs and software.
- Orchestrator / multi‑agent systems.
- RAG agents: AI connected to internal data.
4 Case studies
- HR agent automating employee requests.
- IT agent resolving simple tickets.
- Knowledge assistant agent for centralizing documentation.
5 Use Case Analysis
- Impact / Complexity matrix.
- Identifying quick wins.
- Selection of 3 priority use cases.
6 Scope Definition
- Business objectives.
- Limits and responsibilities of the agent.
- Interactions: user, system, data.
- Risks: hallucinations, security, confidentiality.
7 Design and Tool Selection
- Building a functional diagram.
- Choice of framework (LangChain, AutoGen…) based on:
- Need for autonomy,
- Required integrations,
- Importance of memory,
- Data access (RAG).
- Definition of action logic (tools, APIs, modules).
8 Practical workshop
- Create a workflow prototype for a simple agent:
- Request reception
- Information extraction
- Decision
- Automated action (e.g., email, note, report)
9 Deployment Plan
- Phases: prototype → pilot → production.
- Infrastructure: cloud, API access, key management.
- Change management and adoption.
10 Performance Measurement
- KPIs: success rate, answer quality, speed, user satisfaction.
- Monitoring tools.
- Continuous improvement and maintenance.
See more
📌 Practical information
Our training sessions are offered in Montreal or Quebec City, in person or in a virtual classroom. Dates and locations are specified when you select your session below. If you have any questions, check out our FAQ.
Trainers
Upcoming information