Grigori Melnik, Chief Product Officer, Amperity – Interview Series

Dr. Grigori Melnik is a seasoned technology executive with more than 25 years of experience driving product innovation and growth at companies including Microsoft, Splunk, MongoDB, Tricentis, and Cribl. He has led platform transformations, launched category-defining products, and scaled teams across every stage of growth. Dr. Melnik holds a Ph.D. in Computer Science from the University of Calgary and brings to Amperity a passion for engineering excellence, AI innovation, and building high-impact product organizations.

Amperity is a technology company that offers an AI-powered Customer Data Cloud platform designed to unify fragmented customer data into trusted profiles, identify high-value opportunities, and activate campaigns across all channels. Its solutions focus on identity resolution, data ingestion, and real-time activation, allowing brands to integrate diverse data sources, perform AI-driven analyses, and send targeted audiences to downstream systems. The company emphasizes flexibility by supporting direct connections with major data warehouse platforms and maintains compliance with key security standards such as SOC 2, GDPR, and HIPAA.

You’ve led product and technology strategy at companies like Tricentis, MongoDB, and Codility before joining Amperity. How have these experiences shaped your approach to building and scaling AI-driven platforms like Real-Time Profiles?

By nature, I’m an enthusiast for unsolved problems. At Amperity, we do exactly that. My experiences at previous organizations shaped how I think about scaling platforms while meeting the needs of its users. These lessons include the importance of flexibility, frictionless integration across ecosystems, and strong data governance.

Those lessons have directly shaped our approach to Real-Time Profiles. We purpose-built the capability to end the industry’s oldest compromise—speed vs. accuracy—by unifying historical identity with sub-second streaming in a single, governed architecture. We ensured the platform simplifies customers’ operation models, rather than complicating them. We extended our AI-powered identity foundation to unify real-time and historical data within a single architecture, using one identity graph, one access control layer, and consistent lineage and auditability.

What specific gap or market demand motivated Amperity to develop Real-Time Profiles, and how does it redefine the balance between data speed and accuracy?

Most Customer Data Platforms (CDPs) force teams to pick between acting quickly on shallow, event-only data or acting accurately on profiles that are hours or days out of date. Amperity’s Real-Time Profiles remove that trade-off by continuously joining live signals with the full customer history, so brands can recognize an individual at the exact moment of engagement and respond with context. The result is data that’s complete and current, ready to power in-session personalization and event-triggered journeys with real business impact.

By collapsing batch and streaming into one profile, we move beyond “fast but partial” or “complete but late.” It’s a single, continuously updated customer view that lets marketers and service teams orchestrate next best actions at the speed of intent without sacrificing accuracy.

Can you walk us through the technical underpinnings of unifying historical and streaming data into a single, continuously updated customer profile?

We built a unified data flow with three coordinated layers: ingestion of raw JSON events from any source, continuous processing in a distributed data flow engine, and a live profile store that supports millisecond lookups via our Profile API. Every new click, booking, or loyalty change is reconciled against the same AI-powered identity graph that governs our batch pipelines, meaning no separate identity model, no dual maintenance, no schema drift.

Critically, “identity in motion” ties every event to the durable, stitched profile as it arrives. This enables the instant enrichment of attributes, continuous segmentation, and event-triggered activation using journeys or APIs, while preserving lineage, access controls, and auditability across both analytical and operational workloads.

Many enterprises struggle to operationalize real-time personalization. What are the biggest challenges you see brands facing, and how does Amperity address them?

Consumers now expect every brand interaction to reflect a real-time understanding of their intent, preferences, and history, instantly. Yet, most organizations are constrained by fragmented data systems and delayed insights, which make it difficult for them to respond in the moment. The result is often personalization that feels generic or out of sync with customer needs.

Bridging that gap requires more than faster technology; it demands a unified approach to data and decisioning. At Amperity, we’ve focused on solving that systemic problem by enabling brands to bring together historical knowledge and live context so every interaction can be timely, relevant, and connected to the customer’s full journey. With Real-Time Profiles, brands can power in-session personalization and event-triggered journeys from the same governed source of truth, turning moments like cart abandonment, loyalty tier changes, or on-property check-ins into timely, relevant actions.

How does the integration of AI and machine learning enhance the precision or predictive capabilities of Real-Time Profiles?

AI is the backbone of our identity resolution capabilities, which means that live events are linked to the right person with the right context, such as lifetime value, consent, and loyalty, within milliseconds. That precise stitching elevates every downstream decision: segments recalculate as data changes, profile attributes enrich instantly, and journeys trigger based on the complete customer, not isolated events.

Looking forward, Real-Time Profiles lay the foundation for AI agents to operate with live context by reasoning over evolving profiles, surfacing insights, and autonomously triggering next-best actions across the stack. The combination of AI-resolved identity and streaming context is what unlocks true one-to-one personalization at scale.

From your perspective, how do privacy regulations and data governance factor into building real-time personalization systems?

By extending our existing Customer Data Cloud into streaming, we maintain one governed profile store for both analytical and operational use cases. That coherence helps ensure compliance and auditability while enabling sub-second activations.

Just as important, Real-Time Profiles empower brands to rely on their own first-party data as the trusted foundation for personalization. Every real-time signal is connected to verified, consent-based customer data, so brands can act with confidence that their insights and activations align with privacy expectations and regulatory standards. The same policies and controls that govern historical profiles govern live updates, giving brands immediacy while preserving the trust and strong security posture needed for meaningful, compliant personalization.

With the rise of generative AI, how is Amperity preparing for a future where personalized content could be autonomously generated and delivered in real time?

Generative AI is only as good as the data that fuels it. Real-Time Profiles provide the necessary live, identity-resolved context, so generative systems can tailor content to who the customer is and what they’re doing right now. Our architecture positions AI agents to reason over continuously evolving profiles and trigger next-best actions, bridging insight to activation automatically.

As content generation becomes more autonomous, the gating factor will shift from “can we create it?” to “should we create it now for this customer, given their history and current intent?” Our real-time, identity-aware profiles answer that with precision and governance, enabling safe, relevant, and measurable experiences.

What industries or verticals do you see benefiting most from this technology in the near term, and why?

While all consumer brands benefit from real-time personalization, travel, airlines, retail, and financial services see immediate gains because intent windows are short and context matters. Think upgrades at check-in, re-pricing abandoned bookings, personalized sort order and bundles, or card offers aligned to on-site behavior.

These verticals already operate omnichannel journeys with high stakes for timing, relevance, and service speed and accuracy. By unifying historical identity and live signals, they convert fleeting moments into revenue and loyalty, turning engagement into conversions in real-time.

As Chief Product Officer, how do you measure the success of a release like Real-Time Profiles beyond technical performance — in terms of user adoption or business impact?

We measure success by customer outcomes and adoption, including faster time-to-value for in-session personalization, increased conversion and engagement rates, and improved service metrics across all touchpoints. Our customers have driven results like 2x higher conversions from personalized journeys, millions of new high-value prospects identified outside loyalty programs, and faster, more personal service experiences.

Operationally, I also look for simplification signals like fewer tools to maintain tighter alignment between marketing, data, and service teams. When the same profile powers both analytics and activation without requiring duplicate integrations or data pipelines, you see durable adoption and compounding return on customer data.

Finally, how do you see the role of the CPO evolving as AI becomes central to customer engagement and enterprise growth strategies?

The modern CPO must be the integrator of product, data, and go-to-market outcomes – owning the what and why of the product that turns signals into value. In the world of AI, the CPO must also own the how – how intelligence is embedded into every workflow, interaction, and decision. It also requires  architecting for identity, governance, and real-time action in one coherent solution so teams can deliver experiences at the speed of the customer.

That means blending production management with data science, infrastructure, marketing, customer success, and ethical AI governance to ensure the company’s differentiation doesn’t just come from features, but from learning systems that continuously adapt to users and markets. We are moving from roadmaps to reinforcement loops – our success isn’t about shipping releases but about accelerating cycles of experimentation, learning, and refinement that strengthen both the product and the customer relationship.

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