Former Amazon and Microsoft executive and Glu.ai founder Sangeeta Mudnal speaks to B&T about what really determines how brands get discovered and recommended by AI search.
B&T: For marketers encountering Glu.ai for the first time, how would you explain its role in the martech stack?
Sangeeta Mudnal: Glu.ai is the foundational layer that makes product data trustworthy, structured, and AI‑ready so every downstream tool in your stack performs better. While most martech assumes product data is clean, consistent, and machine‑readable, marketers often struggle with that reality.
Glu sits between your product catalog (PIM/eCommerce platform) and activation tools (CMS, ads, SEO, retail media, AI assistants).
It doesn’t replace existing systems but makes them smarter and more effective. By ensuring AI interprets your products as intended, Glu improves performance across every marketing channel. You can download for free on the Shopify App Store here.
B&T: The name “Glu” suggests connection. What exactly are you trying to “glue together” within the marketing and commerce ecosystem?
SM: Glu.ai is about connecting the parts of the marketing and commerce ecosystem that were never designed to speak to each other but now have to because AI sits in the middle.
With Glu, we’re “gluing together” three critical layers:
- Product truth, what’s actually correct: Your PIM, catalogue, attributes, specs, compliance data. This is the factual source of truth that AI needs but rarely gets in a clean, structured form.
- Brand expression, how you want to show up: Your voice, positioning, messaging, creative intent. This is what marketers care about, but it often gets lost when AI rewrites or interprets content.
- AI interpretation, how machines understand you: Answer engines, shopping agents, LLMs, retail media algorithms. This is the new discovery layer and it’s only as good as the data it’s fed.
B&T: In a crowded martech landscape, differentiation is critical. What makes Glu.ai fundamentally different from existing platforms?
SM: Glu.ai is AI‑native from the ground up, built with AI agents and a data model that treats canonical product facts as first‑class assets rather than an afterthought. Unlike platforms that bolt on AI, Glu delivers rapid PDP optimisation: canonicalising product facts, enriching descriptions for generative prompts, generating FAQs, and fixing schema so product pages are AI‑ready.
With one‑click Shopify publishing, optimised PDPs can go live in about two minutes, enabling brands to iterate quickly and demonstrate GEO lift fast. We pair these capabilities with production monitoring and AI visibility metrics, agency‑ready workflows, and KPI‑first pilots that let brands and agencies operationalise GEO and measure real ROI.
B&T: You’ve spent decades working across AI and eCommerce before launching Glu.ai. What problem did you see emerging in marketing that convinced you it was time to build a new company?
SM: After decades working across AI and eCommerce, the pattern that finally pushed me to start Glu.ai was simple, but impossible to ignore. While marketing is still about persuading a human, success in a future with AI is going to be dependent on being correctly understood by a machine that then persuades the human.
What I saw emerging was a widening gap:
- Brands were still producing human‑oriented marketing content.
- AI systems were making decisions based on structured data they couldn’t trust, parse, or verify.
- And merchants had no visibility into how they were being represented or misrepresented inside AI answers.
That disconnect was creating real business risk. If AI is now the “front door” for product discovery, then inaccurate, incomplete, or unstructured product data doesn’t just hurt SEO it makes a brand invisible in the answer layer. Glu.ai was born from that inflection point.
B&T: What were the biggest lessons from your executive roles at AWS and Microsoft that directly influenced how you built Glu.ai?
SM: One of the biggest lessons from my time at Microsoft and Amazon is that AI must be both scalable and human‑centred. That human‑in‑the‑loop philosophy guides Glu.ai. We build AI agents that are trustworthy, transparent, and comfortable for customers to use. Technically, we rely on three core tenets – clear problem framing, modular architecture, and post‑deployment resilience.
You can prototype an AI agent quickly, but keeping it performant and safe requires a new engineering mindset, most of the work now happens after launch to prevent hallucinations and ensure reliable behaviour. For teams used to deterministic codebases, that shift is a steep but necessary learning curve.
B&T: Australian retailers and eCommerce brands are becoming increasingly sophisticated with their martech stack. How ready are they for platforms like Glu.ai?
SM: Australian retailers and eCommerce brands are directionally ready for platforms like Glu.ai. They have invested heavily in modern martech stacks, personalisation, and are rapidly adopting AI, which means the mindset and infrastructure for a GEO layer already exist. However, most of this adoption is still tactical, focused on efficiency and automation, rather than on strategic influence over how AI systems represent and recommend brands.
There is also a gap in how well teams structure product data and measure visibility across AI-driven surfaces. As a result, while leading brands and agencies are primed to move quickly, Glu.ai’s role is to bridge that last-mile gap and turn existing capabilities into real control and performance in AI-powered discovery. Australia is a strategic growth market for Glu.ai. We are prioritising Australian pilots and agency partnerships to generate early case studies, refine commerce and payments integrations, and prove ROI in real eCommerce environments. For anyone keen to explore it, you can download for free on the Shopify App Store here.
