How Unified Context Turns AI Into Real Enterprise Performance - with Ravi Marwaha of Arango
Enterprise AI agents don't fail because of flawed models—they fail because they lack a live, unified understanding of the business. Ravi Marwaha of Arango argues that AI systems are like newborns: they need nurturing through a real-time context layer that connects data across CRMs, ERPs, support systems, and more. Without it, agents guess instead of knowing, leading to unreliable decisions in high-stakes environments like clinical trials, customer support, and semiconductor design. The breakthrough isn't better models—it's a unified, always-on context layer that enables agents to reason temporally, act with explainability, and deliver measurable ROI. Ravi warns that treating AI as a software project leads to 95% failure rates; instead, organizations must design for context first, integrate existing systems, and treat context as infrastructure—not a one-off task. The future belongs to enterprises that build agents that think like their best employees, grounded in the real-time state of the business.
AI agents fail in production not due to model limitations, but because they lack a real-time, unified context layer that reflects the current state of the business.
Explainability is not an afterthought—it must be engineered into AI systems from the start, enabling decisions to be reconstructed, traced, and audited in seconds.
Building a context layer doesn't require replacing existing data infrastructure; instead, integrate and map entities across current systems like CRMs and ERPs.
Design for always-on context: the system must reflect real-time changes, not just historical data, to support accurate, temporal reasoning.
Prioritize business context over model selection—define the problem first, then build the data and governance needed to solve it.
…and 3 more takeaways available in PodZeus
Introducing the Context Problem
The episode opens with a welcome to the podcast and introduction of Ravi Marwaha, COO and CTO at Arango, a multi-model data platform designed to unify enterprise data for AI.
AI Agents Are Like Newborns—They Need Nurturing
“LLMs are little babies. And then you do have to basically, I'd say, you have to nurture them.”
From Guessing to Knowing: The Shift in Agent Behavior
“There is a big difference between guessing and knowing. Probabilistic answers are okay. Probabilistic decisions and probabilistic actions are not okay.”
The Three Pillars of Effective AI Agents
“Is it relevant? Is it explainable? And is it real time or is it temporal?”
Real-Time Context in Action: Customer Support
Using Zscaler as an example, Ravi shows how a lack of real-time context leads to generic, frustrating support interactions, while a unified context layer enables faster, accurate resolution.
“There is a big difference between guessing and knowing. Probabilistic answers are okay. Probabilistic decisions and probabilistic actions are not okay.”
“It's a liability because if you can't prove it, you cannot do it. Absolutely.”
“Make sure 80 of your energy is spent on that innovation part. Think in terms of agents from the get -go, not as human as I would think.”
Host
Guest
Ravi Marwaha
person
Arango
organization
Shalandi
person
Zscaler
organization
PSI
organization
Emerge AI in Business podcast
media
Lang Pro
product
Gartner Summit
other
Satya Nadella
person
LangChain
product
Modernizing Targeting to Close the Field Execution Gap - with Damion Nero of Daiichi Sankyo
24m • 6/10/2026
The Pricing Shift Reshaping Enterprise AI Spend - with Adam Mansfield of UpperEdge
37m • 6/1/2026
Human-Centered AI Development Strategies for CPG Leaders - with Shaje Ganny of Procter & Gamble
29m • 6/2/2026
How Industrial Service Leaders Are Closing the Knowledge Gap Before It's Too Late with Mike Hughes of Peak International Group
26m • 6/4/2026
How AI Is Reshaping the Way Enterprises Build Software - With Tim Sears of HTEC
21m • 6/5/2026
Start discovering podcast insights today
Start with a 7-day trial and explore a growing catalog of popular podcasts. No credit card required.
No credit card required • 7-day trial • Cancel anytime