The marketing technology world has a new challenger. Kana, a San Francisco-based startup founded by two of the most experienced names in martech, launched from stealth this week with $15 million in seed funding and a platform built around a deceptively simple idea: AI agents for marketers should be flexible enough to actually work in the real world.
The company was co-founded by Tom Chavez (CEO) and Vivek Vaidya (CTO), veterans who have built and sold marketing technology companies before — including Rapt, acquired by Microsoft in 2008, and Krux, acquired by Salesforce in 2016. After those exits, the duo ran startup studio super{set}, where Kana was incubated for nine months before its public debut.
The round was led by Mayfield, with managing partner Navin Chaddha joining Kana’s board of directors.
The Problem With Most AI Marketing Tools
The marketing automation market is crowded. Facebook, Instagram, and TikTok have embedded AI tools directly into their ad interfaces. Microsoft and Google continue expanding their own automation offerings. Content generation startups like Jasper and Copy.ai have captured significant mindshare among digital marketers.
But Chavez and Vaidya argue that most of these tools share a fundamental flaw: they’re rigid. They require lengthy implementation cycles, operate with fixed functionality, and often demand that companies replace existing systems to unlock their value.
That’s the gap Kana is designed to fill.
“Loosely Coupled” Agents: What That Actually Means
Kana’s platform is built around what the founders describe as “loosely coupled” AI agents — specialized modules that handle distinct marketing functions and can be customized, swapped out, or reconfigured in real time as conditions change.
These agents cover a broad range of marketing tasks: data analysis and interpretation from multiple sources, audience segmentation, campaign management and optimization, customer engagement personalization, media planning and budget allocation, and AI chatbot optimization.
The practical upshot is that a marketer could, for example, upload a media brief and have Kana’s agents automatically parse campaign objectives, identify relevant target audiences, and pull in data from inventory systems and market research — all without locking the team into a fixed workflow that can’t be adjusted when priorities shift.
Vaidya framed the positioning clearly: “We operate in a world that enables exploration of a third option with customers: not build, not buy, but build with — build with in a manner that receives proper support. We can move with extraordinary speed that larger corporations simply cannot match. That represents our distinct advantage.”
Synthetic Data: Reducing Costs Without Sacrificing Depth
One of Kana’s more distinctive capabilities is its synthetic data generation engine. Rather than relying solely on third-party data — which is expensive, increasingly regulated, and often full of gaps — Kana can create artificial datasets that mirror real-world marketing patterns while sidestepping the privacy complications tied to actual consumer data.
For marketers, this opens up faster testing cycles, cheaper audience research, and the ability to model campaign scenarios without needing a full third-party data procurement process.
“Synthetic data generation helps companies reduce expenses associated with third-party data procurement, address gaps in existing datasets, and enables marketers to conduct platform tests more rapidly while narrowing strategic approaches,” Chavez noted.
It’s a meaningful differentiator in a landscape where first-party data strategies are becoming essential as cookies disappear and privacy regulations tighten.
Human Oversight Stays Central
One of the more deliberate design choices Kana has made is keeping humans firmly in the loop. Despite the automation capabilities on offer, marketing professionals retain approval authority over the agents’ actions at every stage. They can provide continuous feedback, adjust agent behavior, and override decisions as campaign needs evolve.
This “human-in-the-loop” model reflects a growing tension in AI marketing tools broadly — the desire for automation efficiency on one hand, and the very real need for brand control, compliance, and judgment on the other. Kana is betting that the winning approach isn’t maximum autonomy, but maximum responsiveness to the humans running the campaigns.
A Market Growing at Nearly 30% a Year
The timing of Kana’s launch isn’t accidental. Global spending on AI marketing solutions is projected to exceed $107 billion by 2028, growing at roughly 29% annually — one of the strongest growth trajectories in enterprise software. Investor appetite for specialized AI applications, particularly those with clear commercial use cases beyond general-purpose language models, has intensified accordingly.
Kana plans to deploy its $15 million toward expanding engineering, product development, and go-to-market teams — the classic early-stage playbook for a company that has a working product and is now focused on scaling it.
What Kana Has to Prove
The platform’s real test will come from enterprise marketing teams weighing it against both the big incumbents and a growing wave of AI-native martech startups. Kana’s bet is that flexibility — the ability to integrate with legacy systems, adapt in real time, and avoid the rip-and-replace implementation cycles that slow enterprise adoption — is the feature marketers actually need most right now.
With a founding team that has already done this twice at scale, the credibility is there. Now Kana has to show that its architecture holds up in the complex, messy, deadline-driven reality of actual marketing operations.