Free Lesson

Your AI agent works. Your bill doesn't.

Hosted by Dr Ankur Narang and Kush Khurana

Mon, Aug 3, 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

Where the money actually goes

Measure cost per successful task rather than cost / API call. Identify three items that quietly dominate most agent bill

Model routing that protects quality

Route requests to a lower-cost model first and escalate based on confidence. Understand the decision logic & guardrails

Caching, batching and context discipline

Apply the practical techniques that reduce spend before changing models. Learn how to determine which technique works

Cost-aware evaluation

Track quality and cost in the same evaluation harness so that cheaper does not quietly mean worse.

Why this topic matters

Most agent projects don't die from bad models: They die when someone in finance asks what the system costs per successful outcome and the team has no defensible answer. Inference spend is an architectural decision - not simply a bill that arrives. In this session, participants should be able to model an agent's unit economics, select a routing strategy using evidence and defend the architecture

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; Apparel Group; Yatra; Hike

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

Sign up to join this lesson

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