Free Lesson

Make AI Agents Fail Safely

Hosted by Dr Ankur Narang and Kush Khurana

Tue, Aug 11, 2026

3:00 PM UTC (45 minutes)

Virtual (Zoom)

Free to join

Invite your network

Go deeper with a course

Build Production AI Agents for 10x-100x ROI
Dr Ankur Narang, Kush Khurana, and Dr Aveek Brahmachari
View syllabus

What you'll learn

Classify failures before selecting a response

Learn to distinguish among: transient failures; permanent failures; missing or invalid data; model-quality failure

Add timeouts and bounded retries

See why unlimited retries can make an incident slower, more expensive and harder to diagnose

Build a practical fallback hierarchy

The safest fallback is not always another model. In some cases, the correct production behaviour is to avoid acting.

Preserve workflow state

Understand how explicit states & checkpoints allow an agent to continue from the last successful step vs restarting

Why this topic matters

A production AI agent is connected to models, APIs, databases, enterprise tools and human approval processes. Each dependency can fail independently. Without reliability engineering, one can get: incorrect prices or decisions; duplicate transactions; compliance exposure; poor customer experiences. A production-ready agent must therefore be judged by both its success and its failure behavior

You'll learn from

Dr Ankur Narang

Dr. Ankur Narang brings 30+ yrs exp in AI, Tech across MNCs & many verticals

Dr. Ankur Narang is the Founder of DeepCoreX Labs and brings 30+ years of experience across AI, technology leadership, venture building, enterprise transformation, and applied R&D. He has held senior AI and technology roles across research labs, startups, and large enterprises, and has worked on AI systems, decision science, agentic workflows, and deeptech venture creation.

Kush Khurana

AI & ML leader; Venture Partner, DeepCoreX; Ashoka faculty

Kush Khurana is an AI/ML leader, Venture Partner at DeepCoreX, and Visiting Professor at Ashoka University. He has 12+ years of experience building production AI systems, leading technical teams, and shaping AI-native products. He works with founders and learners to turn frontier AI and deep-tech ideas into practical, scalable systems.

IBM; Oracle; Meta; Apparel Group; Hike

Oracle
Meta
Apparel Group
Hike
Yatra
See all products from DeepCoreX Academy

Sign up to join this lesson

By continuing, you agree to Maven's Terms and Privacy Policy.