Join us for another insightful and hands-on meetup where we take the mystery out of securing AI agent systems! In this session, we’ll explore the tools, practices, and architectural patterns that keep agentic AI infrastructure safe, resilient, and production-ready whether you’re building autonomous agents, multi-agent pipelines, or AI-powered APIs.

Whether you’re a beginner curious about what “agent security” even means, or an experienced ML engineer looking to harden your agentic workflows, this session is for you.

Agenda

Here’s our agenda for the day:

1. Welcome and Introduction  (15 minutes)

• Meet the community and organizers.

  Overview of today’s theme: “Securing Agent Infrastructure: Building Safe, Resilient, and Trustworthy AI Systems.”

• Why agent security matters: From demo to production, the stakes are higher than ever.

2. Understanding the Agent Security Threat Landscape  (45 minutes)

• Prompt Injection & Jailbreaking: How malicious inputs can hijack agent behaviour and how to defend against them.

• Tool & API Abuse: Risks that arise when agents are granted access to external tools, APIs, and databases.

• Data Exfiltration & Privacy Leaks: How sensitive information can leak through agent memory, context windows, and tool outputs.

• Supply Chain Risks: Securing the models, plugins, and third-party integrations your agents depend on.

• Best Practices: How to architect agentic systems with security as a first-class concern from day one.

3. Hands-On: Hardening Your Agent Infrastructure  (60 minutes)

• Authentication & Authorization for Agents: Implementing scoped permissions, API key management, and OAuth flows for agent-facing services.

• Guardrails & Output Validation: Using tools like Guardrails AI, LLM Guard, and custom validators to sanitize and constrain agent outputs.

• Observability & Audit Logging: Tracing agent decisions, logging tool calls, and monitoring for anomalous behaviour in production.

• Sandboxing & Isolation: Running agent-executed code safely using Docker, E2B sandboxes, and containerized tool environments.

• Live Demo & Walkthroughs: Real-world examples of securing MCP servers, LangGraph agents, and multi-agent pipelines.

• Q&A and Community Showcase: Ask questions, share your work, and get feedback from the community.

Gigs

We would love to reach out to you so that you can build for our customers, please fill out this form with details to ensure we have your details:

GIG/HACK DEVELOPER PORTFOLIO FORM

Join community channels

Africa's Talking AI/ML Community:

• Join the Africa's Talking AI/ML Community here

Slack

Join Africa's Talking slack community here

Please follow our Twitter handles to:

• @AfricasTalking

• @ElarianHQ

You can get our videos, recaps, and event interviews on our youtube channels, subscribe to get updates:

• Africa's Talking Youtube Channel

Africa's Talking community allows developers to learn skills for the modern-day African Developer. We are language and framework agnostic. All developers are welcome. This is where Africa's Talking developers community meets to build, learn and exchange knowledge.

We are helping software developers and businesses to bring their ideas to life through easy-to-use APIs easily.

Would you like to partner with us? Kindly contact the Developer Experience Team.