Santanu Dasgupta did not start Meerkats AI from a generic belief that every company needed another AI wrapper. He had spent about 20 years around SaaS go-to-market teams in the US, Europe and India, first as a developer and later in growth and consulting roles. The pattern he saw was repetitive: sales development, marketing and agency teams still stitched together lead data, enrichment, campaigns and follow-up across fragmented tools.

Meerkats AI turned that services-heavy workflow into software. In the Indie Hackers founder interview, Santanu described the product as a digital growth agency in software: agents that capture and enrich relationship data, generate leads, run campaigns and follow up from a chat interface. The current Meerkats site has since sharpened that positioning around performance marketing agents that connect ad platforms, CRM, analytics and revenue data, then optimize toward CAC and revenue instead of cheap leads.

The early distribution matched the customer. Rather than launch to everyone interested in AI, Meerkats targeted agencies that already ran campaigns for clients but were not yet using AI agents. The team combined cold and targeted outreach to agencies in its network, ran educational events online and offline, and used LinkedIn posts to attract users. That made the first wedge practical: sell the outcome to people who already understood the GTM labor cost.

Four weeks after launch, the platform was doing more than $3K MRR. Santanu was careful about the operating reality behind that number: the company also used University of Chicago Polsky Center funding, agency services and Azure, OpenAI and Anthropic credits to keep the lights on. For Startup Founder Stories, that means the clean milestone is early recurring-revenue validation, not proof of mature margin.

The useful lesson is that agentic products get more believable when they own a complete business workflow. Meerkats was not pitched as a toy agent or a faster spreadsheet. It was sold as a way for agencies and startups to move from lead lists to campaigns, follow-up and revenue feedback with fewer handoffs. In a noisy AI category, workflow specificity and distribution through agencies mattered as much as the model stack.