Research & Development

Continual ideation, innovation, & iteration.

Innovate or perish. From our inception in 2005, Aten's foundation has been built on established and proven scientific research — collaborating with NC State, Virginia Tech, Duke, and applying findings from Harvard, Michigan State, NSF, ESA, and the Federation of American Scientists.

The thesis

We remember 10% of what we read,
~30% when we watch a video,
90% when we do it ourselves — even as a simulation.

01 — Cerebrum

Our serious games & simulation platform.

In our early attempts to bridge the global educational divide, we built one of the first platforms in the world enabling students anywhere to learn from teachers anywhere. That work led to development partnerships with NC State and Virginia Tech research teams on the impact of virtual immersive learning — and to enterprise deployments with JPMorgan Chase, Eisai, Biogen Idec, and others.

↳ Hosting

Simulation, VOD & serious-game delivery

A platform built specifically for learning experiences — not a generic LMS retrofitted for video. Web, mobile, and XR delivery from a single source.

↳ Authoring

Evidence-centered design

Every scenario maps to a measurable competency. Branching, constraint, and consequence are designed to elicit the behavior — not test the trivia.

↳ Distribution

Globally portable

Multi-language, multi-region, low-bandwidth fallbacks. The same simulation runs for a New York analyst and a Mumbai trainee.

02 — Cerebellum

Assessment & data tracking, built in.

In 2007 we introduced Cerebellum — our assessment and data tracking platform. It enables embedded assessments and data tracking, including stealth assessments, giving our systems the ability to deliver instant formative feedback as well as in-depth diagnostic and summative analytics with adaptive intelligence.

↳ 01

Stealth & embedded assessment

Measurement happens during the simulation — not as a quiz at the end. Learners reveal their know-how through what they do, not what they recognize on a multiple-choice list.

↳ 02

Adaptive intelligence

Models that observe behavior, surface gaps, and reshape the next moment of learning to the individual. Difficulty, pathing, and reinforcement adapt continuously.

↳ 03

Diagnostic & longitudinal data

Track an individual from hire through years of practice. Surface what's improving, what's plateauing, and what intervention will move the needle next.

03 — Why it works

Beyond multiple choice.

Cerebellum enabled the transition from solely multiple-choice questions — where it's largely content knowledge that's being assessed — to a more accurate measurement of practical application of knowledge using embedded assessments during simulations, serious games, and decision-making scenarios.

Multiple choice

Tests recognition.

  • Did the learner read the slide?
  • Can they pick the right phrase from four?
  • Same questions, same answers, every time
  • No insight into how they'd actually behave
Performance-based

Measures action.

  • What did they do under pressure?
  • How long did they hesitate?
  • Which constraints did they violate?
  • What does their pattern predict on the job?

Want to see Cerebrum in action?

We'll walk you through a working simulation, the data dashboard behind it, and the model that adapts in real time.