turms / Technology
Platform

The intelligence layer.

turms sits between a moving person and a measured outcome, across four layers partners can work with through an agreed integration.

LayerINTELLIGENCE
LatencyREAL_TIME
DecisioningCONTEXT_AWARE
InterfaceAGREED_INTEGRATION
turms intelligence layer

Signals, context, decision, media, outcome.

Signals are raw and continuous. Context is what they mean together. Decision is what should happen. Media is where it happens. Outcome is what it produced, returned to the model.

Signals

What is happening

Movement, environment, vehicle state, time, media availability, real-time inputs.

Context

What it means

Raw signals interpreted as a journey state a media plan can act on.

Decision

What should happen

Movement Media AI (MMAI) coordinates next-best-media, timing and sequence, including defer and suppress.

Media

Where it happens

Audio, in-car, mobile, DOOH, CTV, retail and connected environments.

Outcome

What it produced

Attention, action and incrementality, measured and returned to the model.

Architecture

Four layers between a person and an outcome.

Each layer is addressable on its own. Partners can integrate at the input, consume the decision, or take the measurement feed, without adopting the whole stack.

Physical world

Input

Movement, context, environment, media availability and real-time signals arriving continuously from vehicles, devices and environments.

Context

Interpretation

Signals resolved into journey state. What environment this is, what stage of the journey, what is likely next, and what is reachable from here.

Intelligence

Decisioning

Prediction and sequencing produce the next-best-media interaction. Its timing, its channel and its creative. Or the decision to hold.

Measurement

Outcome

Attention, action and incrementality measured per decision, then returned to the model to improve the next one.

Next-best-media

Six options. turms selects one.

At any moment several media interactions are technically possible. turms selects the one that fits the context and holds the rest back. Doing nothing is a valid decision.

01AudioIn stream or in ad break
02DOOHOn the route ahead
03MobileWhen the car is stationary
04CTVEither side of the journey
05DeferRight message, wrong moment
06SuppressNo interaction is the answer
Data and privacy · In-car advertising (ICA)

In-car: contextual by architecture.

For in-car advertising (ICA), turms decides on the state of the vehicle environment, not the identity of a person. No personal profile is built for in-car decisions. MMAI works differently and may use personal data where permitted, as set out in our Privacy Policy.

Non-personal

Tier-1 signals only

Geolocation, fuel or charge level, driving context and time of day. No PII enters the decision.

Safety-aware

Context gates the channel

The decision layer knows what the vehicle is doing. Interactions that do not belong in motion are not offered.

Regional

Compliance by deployment

Signal handling and retention configured to the regulatory regime the vehicle is operating in.

Auditable

Every decision logged

Each decision, and the reason it was made, is recorded and available for review.

Integration

Built to be integrated, not adopted wholesale.

turms is interoperable with existing programmatic infrastructure. It does not ask a partner to replace a stack that already works.

01

Signal ingestion

Vehicle, device and environment signals arrive through the agreed platform or OEM integration.

02

Decision

Movement Media AI (MMAI) returns the selected environment, timing and creative, or an instruction to hold, through the agreed integration.

03

Delivery

turms stitches into existing streams and ad breaks. No new layer is added to the listening experience.

04

Measurement feed

Outcome data returns in the agreed reporting format and reconciles against existing reporting.

Integrate the decision layer.

Talk to turms about integration options for platform and programmatic partners.