Chetan Alsisaria, CEO and Co-Founder, Polestar Analytics – Interview Series

Chetan Alsisaria, CEO and Co-Founder is an enterprise technology leader specializing in data, analytics, and AI-driven transformation. After early roles at Deloitte, PwC, and EY, he co-founded Polestar Analytics in 2012 and has since scaled it into a global AI and data company. He leads sales, strategic alliances, service development, and technical delivery, while also serving as Chair of CAIO Circle, a community for AI leaders focused on advancing responsible and practical AI adoption. In addition, he co-founded Xumane Equity, reflecting his broader focus on innovation across enterprise systems and platforms.

Polestar Analytics is a global AI and data convergence company that helps enterprises turn fragmented data into actionable insights through its proprietary 1Platform. By unifying data engineering, analytics, AI, and business workflows into a single ecosystem, the company enables organizations to improve decision-making, optimize operations, and scale AI adoption. With a strong focus on industry-specific use cases and measurable outcomes, Polestar has evolved from a consulting-led firm into a platform-driven business delivering simplified, intelligence-driven solutions at enterprise scale.

You co-founded Polestar Analytics in 2012 after roles at Deloitte, PwC, and EY. What gap in enterprise data and analytics did you see at the time, and how has that original vision evolved into today’s AI-driven 1Platform?

You know, when Ajay, Amit, and I started Polestar Analytics back in 2012, the irony was striking; companies were drowning in data but starving for decisions. Everyone was collecting everything, but the gap between having data and doing something meaningful with it was enormous. That’s the problem we set out to solve.

Fast forward to today, and honestly, the problem hasn’t gone away, it’s just shapeshifted. The volume is bigger, the stakes are higher, and now you’ve got agents in the mix alongside humans who need to make sense of it all. That’s actually made our original vision clearer, not murkier.

Our motto, Data to Outcomes, Simplified!, is really the thread that connects 2012 to today. We started with data deliverability; getting the right data to the right people at the right time. Then AI came in and amplified what was possible. Now with 1Platform, we’re pushing further; simplifying and maximizing outcomes not just for humans but for the agents that are increasingly making or informing decisions.

What’s exciting is how our ecosystem has matured to support this. Our deep integrations with Microsoft, Databricks, and Anaplan help bring data, business use cases, and planning together. And 1Platform sits across all of this, native to these environments, not bolted on.

So, the evolution really looks like this: data deliverability → AI-augmented insights → simplified, agent-ready outcomes. The pursuit of excellence is the same. The speed at which we can get there is what has changed exponentially.

Polestar Analytics positions itself as a data and AI convergence company. What does convergence actually look like in practice for large enterprises dealing with fragmented systems and siloed data?

Most of the time, fragmentation is not a technology problem, it’s a people and process problem. It’s like you have a finance team running on Anaplan, an ops team living in Excel, a data engineering team building pipelines on Azure, and everyone pulling in different directions with different definitions of the same metric. No amount of AI fixes a trust problem if the foundation is broken.

So, when we talk about convergence, we mean solving that foundation first. Before you can layer intelligence on top, you need data that’s clean, governed, and critically accessible. Not just to your analysts, but increasingly to your agents too.

In practice, convergence with 1Platform looks like this: we’re not ripping and replacing what enterprises have built. We go native into the environments they already live in, such as Databricks, Microsoft, and Anaplan, and we stitch the data and intelligence layer across them. Your planning data in Anaplan talks to your operational data in Databricks, and your Microsoft ecosystem is where decisions actually surface for people and agents to act on.

The magic isn’t in any one of those integrations. It’s in the connective layer of 1Platform that makes everything feel like one coherent system instead of disconnected tools. That’s what convergence looks like in practice, a deliberate simplification until the complexity becomes invisible to the business.

Your proprietary 1Platform aims to unify data, AI, and workflows into a single system. How does this approach differ from traditional BI stacks or modern data platforms like Databricks or Snowflake?

Databricks and Snowflake are powerful platforms, we’re not competing with them, we’re built on top of them. That distinction matters. Databricks gives you infrastructure and compute. We sit above that and ask a different question: now what?

Traditional BI stacks got many things right for their time, but business intelligence has evolved. Today, business users need more than dashboards. You can have a beautifully designed dashboard with fifteen charts, and someone still has to interpret what it means and decide what to do next. That gap between insight and action is exactly where 1Platform operates.

1Platform is not static, it continuously evolves. It doesn’t just answer the questions you bring; it surfaces questions you didn’t even think to ask. That changes the relationship between business users and their data.

We’ve built low-code and no-code interfaces on top of Databricks and Azure that allow pipelines to be spun up in seconds. Tasks that once took days for data engineers can now be triggered by business users. On top of that, Agenthood AI enables users to create and orchestrate agents through simple drag-and-drop interfaces without deep technical expertise.

But the real difference is the end-user experience. Instead of jumping across multiple dashboards, users receive natural language insights, contextual recommendations, and agent-driven narratives. KPIs don’t just sit on a screen; agents actively monitor them, flag what matters, and explain why. The differentiation is not the data platform itself, but everything that happens after the data is ready.

Many enterprises remain stuck in what is often called AI pilot purgatory. What are the biggest structural or organizational barriers preventing AI from reaching production at scale?

I call it the pilot graveyard because most projects don’t just stall, they quietly die. The biggest barriers aren’t technical, they’re organizational. People, processes, and data.

Change management is consistently underestimated. When you redesign how people work, you challenge how they derive value. Organizations that succeed are those where leadership makes AI proficiency visibly important. When upskilling is rewarded and process redesign is supported structurally, adoption accelerates.

Then there’s the J-curve problem. AI investments often dip before they rise. Many organizations expect ROI within 90 days, don’t see it, and abandon the effort. The ones that succeed commit to the full curve.

Data readiness is another critical factor. Poor data leads to confidently wrong decisions. Until the data foundation is trusted, AI at scale becomes a liability rather than an asset.

Finally, use case discipline matters. Instead of trying to do everything with AI, organizations need to focus on use cases that move real business metrics, prove them, and then scale.

At Polestar Analytics, bringing data, AI, and workflows into one place accelerates both problem discovery and opportunity identification, turning convergence into a catalyst for change.

Agentic AI is becoming a major theme across the industry. How is Polestar Analytics thinking about AI agents within enterprise workflows, and what real-world use cases are gaining traction?

For us, agents must be embedded within both the data layer and business workflows to create real value. A pricing agent, for example, isn’t just an LLM sitting on a dashboard; it’s integrated into the data infrastructure, understands context, and supports real decision-making.

Across our 100+ agents, some act as assistants while others are fully automated. The strongest traction is in revenue growth management, including pricing, promotions, and media mix, where decisions are frequent and data-intensive.

On the engineering side, pipeline monitoring and error resolution agents are already in use. FinOps agents have reduced unutilized cloud costs by 35 percent. There is also strong adoption in wealth management.

We support both Microsoft and Databricks ecosystems and provide a custom agent-building platform. Not every agent needs to be LLM-based; architecture should match the use case to balance scale and cost.

Governance is essential. With agents interacting with financial systems or customer data, strong guardrails and human oversight are built in to ensure reliability at enterprise scale.

With your recent funding, you are doubling down on IP development. How important is owning proprietary platforms compared to building on top of existing ecosystems in today’s AI landscape?

We are deeply integrated with platforms like Databricks, Microsoft, and Anaplan, which provide infrastructure and scale. We are not trying to replace them.

Our focus is on owning the intelligence layer on top. Proprietary IP allows us to control the experience, embed domain knowledge, and deliver consistent value at scale.

Our differentiation comes from industry expertise. Whether it’s PromoPulse AI for revenue growth management or WealthPulse for financial services, the value lies in understanding real-world use cases and decisions.

Proprietary IP, for us, is the codification of that expertise. It’s what makes the platform defensible and genuinely useful.

You have worked closely with Fortune 1000 companies. How are expectations around AI ROI changing as executives demand measurable outcomes instead of experimentation?

The shift is real, but experimentation hasn’t disappeared. It’s just expected to move faster and connect to tangible outcomes.

Executives now evaluate ROI through broader lenses such as decision velocity, customer loyalty, innovation capacity, and resilience.

AI ROI is no longer owned by a single leader. The CTO focuses on infrastructure and data, the CFO looks at financial impact, and the COO emphasizes operational efficiency.

Organizations that succeed align these perspectives early and commit to long-term outcomes.

Your expansion strategy includes North America and Europe. What differences are you seeing in AI adoption maturity and enterprise readiness across these regions?

The difference is more about mindset than capability.

North America prioritizes speed and experimentation, driven by competitive pressure.

Europe emphasizes governance and ethical AI from the outset. However, this does not mean slower adoption. Enterprises are balancing structure with acceleration.

Both regions are converging toward scalable, responsible AI embedded into core operations.

You recently founded the CAIO Circle to bring AI leaders together. What are the most urgent conversations happening among Chief AI Officers right now, especially around governance and ethics?

CAIO Circle was created to give AI leaders a space for open discussion.

The central challenge is balancing speed with long-term risk. Trust and explainability are major concerns, especially as AI systems influence critical decisions.

Governance is shifting from policy documents to embedded operational practices. At the same time, many organizations still lack execution models to deliver on their AI strategies.

The most valuable insights often come from candid peer discussions rather than formal presentations.

Looking ahead three to five years, do you expect enterprise AI to consolidate into unified platforms like yours, or remain a fragmented ecosystem of tools and vendors? Where does Polestar Analytics aim to position itself in that future?

Fragmentation will likely continue. Major platforms like Databricks, Microsoft, Salesforce, and Anaplan will remain central players.

What enterprises need is a unifying layer that connects data, intelligence, and workflows into something actionable. That is the role 1Platform is designed to play.

The future will favor platforms with deep vertical expertise. Generic horizontal solutions will struggle to differentiate.

Real value will come from understanding industry-specific needs and embedding intelligence directly into decision-making workflows.

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