Sandeep Menon, CEO & Co-Founder at Auxia – Interview Series

Sandeep Menon, CEO & Co-Founder at Auxia, brings over two decades of global technology and marketing leadership to the role. Before launching Auxia in 2022, he spent more than nine years at Google, where he served as Vice President of Marketing for Payments and led initiatives such as the Next Billion Users program, focused on digital inclusion in emerging markets. He also held senior marketing leadership roles across Android, Chrome, ChromeOS, and Google Play.

Auxia is an AI-driven marketing platform that empowers enterprise teams to orchestrate personalized, 1:1 customer journeys at scale. Instead of relying on rigid, rules-based campaigns, Auxia lets AI agents test hypotheses across all touchpoints—email, web, app, offers—and dynamically adapt to real-time preferences and behavior. It integrates seamlessly with first-party data sources, automates ML pipelines, and continuously optimizes messaging, timing, and sequencing to maximize engagement and lifetime value.

You founded Auxia after a successful career at Google, where you led global marketing for products used by billions. What specific gap or pain point did you and your co-founders see in the market that led to the creation of Auxia?

At Google, I saw firsthand how powerful true personalization can be when you have the right infrastructure and AI capabilities. But when I looked at the broader market, I saw companies struggling with fragmented marketing stacks, often managing 12 to 14 different point solutions that didn’t communicate with one another. They were collecting massive amounts of customer data but couldn’t translate it into meaningful, real-time personalization.

The fundamental gap was that existing platforms were built for the pre-AI era. They relied on static rules and basic segmentation, when what businesses really needed were intelligent, adaptive systems capable of making real-time decisions about every customer interaction. We saw an opportunity to bring the same level of AI-driven personalization that powers companies like Google and Meta to businesses of all sizes, without requiring them to build massive internal data science teams.

Auxia’s founding team includes former leaders from Google, Meta, and Lyft. How did your collective experiences at these tech giants shape the architecture and ethos of Auxia?

We’ve all experienced the challenges of personalization at scale. On the Google Pay team specifically, we were in a privileged position because we had early exposure to transformative technologies in their infancy, with transformer models being a prime example. My co-founders, who came from Meta and Lyft, built systems that pioneered ways to take the vast amount of customer data available and make it dramatically more useful, like powering real-time recommendations and decisions for millions of users every second.

We all saw firsthand how AI could profoundly improve customer experiences while also addressing some of the toughest challenges faced by our marketing teams. At the same time, the broader market was shifting in a way that made this kind of work possible outside Google. With the rise of the modern data stack and platforms like Snowflake, BigQuery, and Databricks, enterprises had spent the past 5-10 years consolidating their data into one centralized place ready for activation.

We recognized a unique opportunity to take the learnings we had developed inside Google and democratize them for enterprises everywhere. These companies were just waking up to the realization that they were sitting on a mountain of valuable data and with the right technology, it could be unlocked to drive growth and better customer engagement.

Most personalization tools rely on rules-based systems and simple segments. Auxia instead uses synchronized AI agents. Can you walk us through how these agents collaborate and evolve over time to personalize each customer journey?

Most tools on the market today still rely on rigid rules and static segments, they’re reactive, not intelligent. At Auxia, we’ve built a system of synchronized AI agents that work together to personalize every touchpoint in the customer journey in real time.

Each agent plays a specialized role. Decision agents determine the best action to take for a user based on all of their previous data and preferences; for example, whether to upsell a customer to a new credit card or drive them to open a new savings account.  Analyst agents function like a built-in data science team, analyzing what has worked well in current campaigns to highlight impact but also opportunities for improvement in a chat-based interface. Lastly, Auxia’s Content agents leverage all of the data and insights around what has worked best to proactively surface new messaging or creative variations for marketing teams to approve.

What makes these agents powerful is that they don’t operate in silos. They collaborate continuously, learning from every interaction and adapting based on what’s working. Marketers set high-level goals, and the agents handle the complexity. They’re self-optimizing systems that evolve daily to drive better outcomes at scale.

Auxia processes 2.6 billion events daily and handles 6,500 queries per second. What infrastructure innovations enabled this kind of real-time scalability so early in the company’s lifecycle?

From day one, we knew scale had to be non-negotiable. Hyperpersonalization only works if you can make decisions in milliseconds, using fresh, contextual signals. We learned a lot of these lessons from our time at Google and it took over a year to build the infrastructure to support that level of real-time orchestration.

Our architecture is cloud-native, event-driven, and optimized for high-throughput streaming, so we’re standing on the shoulders of technology waves before us like the move to the modern data stack. The way the marketing tech landscape has evolved towards open ecosystems have allowed us to move much more quickly, enabling us to integrate directly with modern cloud data warehouses, CRMs, and other marketing platforms. This has allowed us to build for scale for day one and also has helped us rely on infrastructure that can process massive volumes of first-party customer signals without introducing latency.

What also sets us apart is that we’re AI-native from the get-go. We Auxia designed it to support large-scale experimentation, running thousands of simultaneous hypotheses and continuously optimizing outcomes without human intervention, but we also are leveraging the latest technology in the industry for feature creation, model deployment, and LLM inference.

How does Auxia’s model experimentation framework differ from traditional A/B testing, and what have been the most surprising insights uncovered using your ML-driven approach?

Traditional A/B testing is incredibly limiting, you can only test a few variants at a time, and it often takes weeks or months to reach statistically significant results. At Auxia, we’ve reimagined experimentation as a continuous, intelligent process. Our model-driven framework enables marketers to run many different self-optimizing models and test hundreds of variations in parallel. Using techniques like reinforcement learning, Auxia’s decisioning system makes real-time decisions based on live data.

Let’s take the example of an offers and rewards promotion campaign over email. Previously, marketers might test 2-3 different offer variations, like 5% off, 10% off, and 20% off, divide the audience into mutually exclusive groups, launch, and compare the results across a variety of metrics to see which variation worked best, on average, for the entire group. That winning variation would then be deployed to everyone after the experiment period concludes. There are a few challenges with this approach. First, most experiments fail, so it often takes teams weeks or months to identify what’s working well. Additionally, they are super manual and time consuming to set up, which limits your team’s velocity and how quickly you’re able to deliver impact.

With Auxia, marketers first set a high level goal, like driving purchases. From there, your team would set up hundreds of different variations in the system, which a team can dynamically generate with our Content Agent, ingest from your CMS, or manually define. Each of these variations would vary in terms of the offer construct (e.g. 5% off vs. BOGO), amount, content, and potentially even channel (e.g. email vs. SMS). For each individual user, Auxia’s Decision Agent will rank, score, and predict, out of all the 100s of variations that are available, what is the optimal variation for each person to drive your core objective.

Auxia’s platform is designed to remove the dependency on large internal data science teams. Have you seen traditional marketing teams successfully take on more technical roles with your tools?

One of the core reasons we started Auxia was to empower marketing teams to personalize both their marketing and product experiences without relying on engineers or data scientists. Within just a few weeks of onboarding, marketers using Auxia are able to launch new campaigns, pull data, and interpret results completely on their own. As the world moves toward a more agentic future, we believe marketers won’t need to become more technical or build new technical skills. Instead, the responsibilities traditionally handled by engineers or analysts will be democratized through intelligent agents that augment the marketer’s workflow and overall experience. With our Analyst Agent, we’ve already succeeded in abstracting away much of that complexity of traditional data scientist/analyst work, and we’re excited to continue pushing those capabilities forward.

As someone who led marketing for the Next Billion Users initiative at Google, what parallels do you see between inclusive tech adoption and agentic AI in marketing today?

When I led marketing for the Next Billion Users initiative at Google, we recognized that people in emerging markets interacted with products in fundamentally different ways than users in developed markets. Because that represented such a massive growth opportunity, it became vital to build specifically for those users — simplifying interfaces, abstracting away complexity, and ensuring the experience was accessible and empowering.

I see a very similar trend unfolding with agentic AI in marketing. Just as many of those new internet users leapfrogged directly into mobile-first experiences without needing prior digital literacy, marketers today don’t need to go back and learn SQL or data science in order to unlock personalization. Agents bridge those skill gaps and make sophisticated capabilities immediately usable. At the same time, the way people interact with AI — moving beyond chat-based interfaces to more intuitive, context-aware agents — is forcing developers to rethink the UI of the future. The emphasis, as with NBUs, is on accessibility, simplicity, and empowerment: abstracting away the technical complexity so the end experience feels natural and impactful.

What’s the biggest misconception companies have about AI personalization today, and how do you help them overcome it?

The biggest misconception companies have about AI personalization is that adopting it means giving up total control or diminishing the strategic impact of marketing teams. AI excels at processing complexity, like analyzing massive datasets, spotting patterns that humans can’t see, and surfacing signals that matter in real time. What it can’t do is provide the context, empathy, and strategic judgment required to design meaningful customer experiences. That’s where humans shine: setting the vision, defining brand guardrails, and understanding when a personalization choice could feel intrusive or off-brand.

If we’re speaking candidly, using today’s technology, humans can be effective without AI, but AI can’t be effective without humans. The real breakthrough comes when you augment your teams with AI to handle the scale and complexity of data-driven decisioning, while empowering teams to focus on strategy, creativity, and empathy. That’s when personalization moves from being a buzzword to something that genuinely creates value for both customers and the business.

With backing from over 50 industry leaders and a $23.5M raise, what are the key areas you’re prioritizing for the next 12–18 months of product and team growth?

We’re prioritizing three areas of growth over the next 12–18 months. First, we’re committed to delivering exceptional value to our customers by continuing to elevate their experience. This has always been a top priority for our team and we want to ensure we’re delivering exceptional ROI for all the companies that have put their trust in us and our team. Second, we’re expanding our AI agent capabilities to support enhancing the full marketer workflow, helping teams create content, orchestrate personalized experiences, and surface actionable insights on what’s driving impact. Finally, we’re scaling our go-to-market engine. With strong product–market fit in the enterprise, the next step is to expand our reach by growing our sales and customer success teams. This allows us to bring AI-driven personalization to more businesses, while ensuring new customers benefit from the same high-touch support and differentiated experience that has fueled our early success.

What excites you most about the next frontier of agentic AI, not just in marketing, but in other enterprise applications as well?

What really gets me excited is seeing the pace of adoption, seeing that adoption have real impact, and also seeing the market recognize that agentic AI’s real value is in augmentation, not necessarily replacement. Too many conversations today focus on AI as a substitute for human roles, like all the billboards you might see on the 101 heading into San Francisco about companies replacing SDRs (or even entire functions). That narrative is going to grab headlines, but for me it misses the bigger opportunity: empowering people to do their best work.

Humans bring context, empathy, creativity, and judgment that AI simply cannot replicate. Where agentic AI shines is in handling the complexity and scale that bogs teams down. We’ve talked a ton about the applicability to marketing workflows, like surfacing insights from massive datasets, automating repetitive tasks, and orchestrating processes across systems in real time. When those capabilities are paired with human strengths, the result is a step-change in what organizations can achieve.

In marketing, that means freeing teams from endless execution so they can focus on strategy, storytelling, and customer empathy. But this same pattern applies across all functions in the enterprise: in sales, agents can qualify and prep opportunities so humans can spend more time building relationships; in customer success, agents can flag risk and opportunity signals so humans can deepen partnerships; in engineering, agents can accelerate development by automating testing, debugging, and code generation, so teams can focus on solving complex architectural problems and driving innovation.

All of these opportunities will create trillions of dollars in value and I love seeing the companies that are pushing the envelope to shape how the work people do will evolve over time.

Imagine finance teams with AI agents optimizing budgets in real time, HR teams using agents to personalize employee engagement, or customer support agents that move from reactive to proactive.

We’re just getting started. The future isn’t about AI replacing humans, it’s about humans and AI agents working together to make decisions that are faster, smarter, and more empathetic. That’s the future we’re building at Auxia.

Thank you for the great interview, readers who wish to learn more should visit Auxia.