Ajay Khanna remembers the attraction plainly. He watched a podcast conversation between Yugabyte co-founder Karthik Ranganathan and data analyst Sanjeev Mohan, and something clicked. Here was a difficult product in an old, enormous market, explained with enough life that he wanted into the room. A job search can begin with a recruiter. His began, in spirit, with a good episode and a stronger question: what could this company become if more people understood it?
In February 2025, Khanna joined Yugabyte as chief marketing officer. The company builds YugabyteDB, an open-source, PostgreSQL-compatible distributed database designed for applications that cannot politely stop when a cloud zone has a bad afternoon. His assignment spans a familiar divide. On one side sit consensus replication, multi-region deployment, transactional guarantees and migration paths. On the other sit developers and executives with urgent, perfectly human concerns: keep the application available, avoid lock-in, control cost, and do not make modernization feel like open-heart surgery.
This is where Khanna has spent much of his career - in the narrow, busy corridor between complex machinery and a market deciding what to call it. He is not a database engineer. His training, however, began with engineering, and his first professional steps at Amdocs were in product design and user experience before he moved into product management. Marketing came later. The sequence gave him a useful habit: begin with how a thing works and how a person encounters it, then decide what deserves to be said.
A career built in the middle
Khanna earned an engineering degree from Thapar Institute of Engineering and Technology and an MBA in marketing and finance from Santa Clara University. The pairing sounds almost suspiciously tidy for a future enterprise-software CMO: enough technical vocabulary to interrogate a product, enough commercial training to interrogate a market. His work then carried him through Amdocs, Progress Software, KANA, Oracle and Veeva Systems, followed by longer building chapters at Reltio and Explorium.
At Reltio, Khanna helped build the marketing function during the company's rise in modern data management. The product promised to connect scattered customer information and make it usable across an enterprise. His public writing from that period avoided treating data as a trophy to be collected. The point was action: understand an account, improve a decision, reduce the distance between a signal and a useful response. A former colleague, Ramon Chen, later credited Khanna's contributions as a major part of Reltio's early success.
Another former teammate remembered something smaller and perhaps more revealing: lunchtime walks filled with advice, growing responsibilities handed over as trust accumulated, and a manager who stayed cheerful when conditions were not. Startup mythology loves the heroic sprint. Actual organizations are built in quieter increments - a useful conversation, an assignment that stretches someone, a leader who does not export panic. Khanna's reputation among colleagues suggests he understood that marketing capacity is grown in people before it appears on a dashboard.
The fourth first time
When Khanna appeared on a marketing podcast in 2022, then as Explorium's CMO, he described it as the fourth time he was building a marketing organization. The host teased him about repeatedly choosing startup difficulty over corporate stability. Khanna laughed, accepted “glutton for punishment” as a fair opening joke, and then supplied the real reason. He liked the pace, the learning and the passion of close teams trying to build something before certainty arrived.
There is no fixed playbook hiding in that repetition. Khanna's more useful observation is that a startup's positioning is never finished. He has argued that a leadership team should keep asking what its story means, revisiting the question every six months and considering a larger repositioning as the product, customer and market change. At one previous company, he recalled, the work happened two and a half times. The fraction is funny because it is true: brands rarely change on a ceremonial schedule. They evolve in messy releases.
This iterative instinct is what marketers can steal. A category statement is not scripture. It is a working interface between what the product can do, what buyers need and what competitors have taught everyone to expect. Leave it unattended and once-sharp language becomes wallpaper. Change it too often and nobody remembers you. Khanna's cadence offers a middle path: inspect frequently, rebuild deliberately.
Audit the company story like a product. What changed in the buyer, the market and the thing you sell? Keep the language that still carries weight. Retire the rest.
Selling the calm day
Yugabyte gives that practice unusually consequential material. Khanna explains two broad uses for YugabyteDB. It can serve as a PostgreSQL-compatible transactional database for businesses such as commerce platforms and mobile carriers. It can also become the distributed core of a cloud-native system, spreading work across infrastructure and geography. The technical competitors differ by use case. The customer anxiety does not. They need a modern system without surrendering availability, portability or control.
At KubeCon North America in 2025, Khanna framed the company's challenge as developer awareness. When developers outgrow a monolithic transactional database and need higher availability or multi-region operation, he wants YugabyteDB to enter the consideration set. Open source matters in that argument. So does the ability to deploy in public clouds or on premises. The marketing task is not to make every developer admire distributed systems. It is to make the option visible at the exact moment an old architecture becomes a business constraint.
His sharpest line from another KubeCon conversation was, “Don't let success be the reason for your failure.” It compresses an architectural problem into an operator's nightmare: an application attracts demand, the demand exposes a bottleneck, and popularity produces the outage. The phrase works because it moves the database out of the basement. Scale is no longer an abstract virtue. It is the capacity to survive one's own good news.
That distinction also keeps the story honest. Resilience is not exciting in the ordinary sense. It resembles plumbing, brakes and a well-run train timetable: most appreciated when absent. Khanna's job is to sell the calm day, when a zone fails and customers continue tapping, buying, messaging or reconciling accounts. The feature list earns credibility. The uninterrupted Tuesday earns attention.
The data thread finds AI
Across Khanna's career, the nouns have changed while the underlying concern has remained remarkably stable. At Reltio, it was connected customer data. At Explorium, where he became CMO in 2021, it was finding external signals that could improve analytics and machine-learning models. At Yugabyte, it is the operational data foundation beneath global applications. The common question is whether organizations can turn scattered information into reliable action.
That thread now runs into agentic AI. In 2026, Khanna demonstrated a campaign workflow built with seven agents using Lyzr AI and Yugabyte Meko, a shared-context data layer. Demand generation, product marketing and design could work against an evolving pool of conversations, memory and knowledge. It was a fitting experiment for a marketer who has spent years arguing that data becomes valuable when it is current, connected and available to the people doing the work.
The interesting part is not that a CMO used agents. Novelty is cheap and has an excellent press agent. The interesting part is organizational: the system was designed to learn as teams designed campaigns together. It treated context less like a file to be periodically rebuilt and more like shared working memory. Khanna said he planned to explain how the team built it. The builder, after all, is happiest near the scaffolding.
An appetite for unfinished things
Khanna's public persona is warm, energetic and practical. He talks about speed without pretending resources are infinite. He values peer groups because senior marketing can otherwise become an isolating act of performed certainty. He returns to trust - between marketing and sales, leaders and teams, companies and customers - because good data and polished positioning are useless when nobody believes the person presenting them.
His aspiration at Yugabyte is specific enough to measure and broad enough to take years: help more developers recognize when a distributed, open-source PostgreSQL database fits the move to cloud-native applications. The work involves events, technical education, customer examples and a constantly tuned story. It also involves resisting the temptation to declare the category settled merely because a slide deck says so.
There is an elegant loop in his arrival. A thoughtful technical conversation drew him to Yugabyte; now he is responsible for creating the conversations that draw in everybody else. The podcast was recruiting, product education and category marketing at once. It showed that difficult ideas do not need to be flattened. They need a patient translator who knows which complexity is essential and which is merely showing off.
After more than two decades, Khanna still chooses markets with moving definitions and teams that must be assembled while the work is underway. He calls himself a builder at heart. The phrase explains the job better than the title does. A chief marketing officer can inherit a brand. A builder notices the loose beams, invites people to the workbench, and keeps asking whether the structure still fits the life happening inside it.