ThinkingAI has announced the Agentic Engine, an autonomous growth engine for consumer and gaming companies. Its AI agents instrument the data, diagnose the cause, decide the response and carry it out, from tracking plans to segments to live campaigns. The engine runs on the customer’s own infrastructure and models, including self-hosted, on-premises and virtual private cloud deployments.

“Everyone’s models are smart enough now,” said Chris Han, co-founder of ThinkingAI. “What customers keep asking is whether they can trust an agent to touch live revenue, and closing that gap is the whole game. That’s where the agents do the work, in your data, on your infrastructure, with a person still accountable for every call.”

Closing the Loop on Four Bottlenecks

Analytics agents on competing platforms surface a finding and hand it back to a person. The Agentic Engine takes it the rest of the way, turning it into a product change: an updated tracking plan, a rebuilt segment, a launched campaign. ThinkingAI calls the category agentic growth.

ThinkingAI built the Agentic Engine around four bottlenecks it sees across the category. Teams wait weeks to act on what the data already told them, and revenue leaks in the gap. Answers come from a small pool of analysts and engineers, so a junior operator queues up behind them. Pricing tied to events and monthly tracked users charges a company more for closely monitoring its business, which teaches teams to measure less. And the stack pulls apart: data warehouses, business intelligence tools, analytics platforms and engagement software that drift out of step. The engine answers the first two by doing the work itself, the third by not charging per event, and the last by bringing analysis and engagement onto one platform.

Product Details

The Agentic Engine includes:

  • Agents for data tracking, analysis, engagement and experimentation that work on the same problem together
  • Knowledge Base, a schema-first entity graph that turns a company’s existing documentation and reports into one source of truth, with versioned change tracking
  • ae-cli, a command-line interface that lets a company’s other AI agents and developer tools call the engine directly, without going through its screens
  • Human-in-the-loop guardrails, audit trails and role permissions that trace every autonomous action to the person accountable for it
  • Deployment on managed cloud, self-hosted, on-premises or a virtual private cloud, with customers free to run their own models

Companies covered by the European Union’s General Data Protection Regulation, or facing an enterprise security review, can’t hand over raw player and customer data to an open cloud system, so the engine runs within their own environment. ThinkingAI reports that data-tracking implementation drops from weeks to hours and that a junior operator gets the same answer a senior analyst would have produced.

The engine was built against gaming workloads, where retention and monetization are decided moment to moment. ThinkingAI has spent 10 years working with behavioral data across more than 1,500 enterprises and 8,000 applications, including Sega, Krafton, Habby and Century Games. It is now selling that work to consumer companies outside the gaming industry.

“Cost savings are the easy part,” said Brandon Nader, vice president of marketing at ThinkingAI. “The number I care about is how long it takes to get from noticing something in the data to changing something in the product. Most teams still measure that in weeks. That’s where the money goes.”

Availability

The Agentic Engine is an enterprise software subscription with no per-event charge. Smaller teams can deploy it through cloud marketplaces, while large deployments that carry data-sovereignty and scale requirements go through direct enterprise sales.

About ThinkingAI

ThinkingAI, formerly ThinkingData, builds autonomous growth software for consumer and gaming companies. For its first decade, the company helped enterprises understand their behavioral data. It now builds agents that sense, plan and act on that data inside the customer’s own infrastructure.

Media gallery

About The Author
Business Wire