AHOY Labs is the research department of AHOY Technology. We de-risk deep technology and ship it as SDKs that power AHOY Machine Studio and partner products.

Can the SDKs run on our own hardware?

Yes. Every lab SDK is cloud-agnostic and can run on the edge, fully on premises, in the cloud, or across sites.

How do we get access?

Try the live playgrounds in your browser, then license the live SDKs through AHOY AMS. For a full solution, Build with AHOY. For evaluations, research partnerships or integration questions, write to research@trouve.works.

Are Tokenization and GenAI available?

Both are in research. Early partner conversations are open, and they are shaped with the teams that will use them first.

A live point-cloud scan of a city at night, with vehicles tracked by detection boxes

R&D × AHOY Technology × Physical AI

Finding efficiencies at every curve.

One city, five SDKs. Scroll to see howSee how AHOY Labs research helps machines see, understand and act in the physical world.

Explore the SDKs Our research

Perception SDK · Live

See what matters, frame by frame.

Cameras already on the street become sensors. People, vehicles and assets are detected and tracked, and anything unusual is flagged.

Perception SDK

Voice SDK · Live

Let people talk to the city.

Voice agents clean real-world audio, understand the request and answer in real time, grounded in your own data.

Voice SDK

Data Intelligence SDK · Live

Give every event context.

A knowledge graph above the city links what is seen and heard to what is known, and answers with sources.

Data Intelligence SDK

Tokenization SDK · In research

Move data without exposing it.

Secure tokens carry what the system needs to the edge, while the raw data stays in its vault.

Tokenization SDK

GenAI SDK · In research

Generate for every fan.

Multimodal generation, from expressive dubbing to personalised visuals, built from work the lab has already shipped.

GenAI SDK

AHOY Machine Studio

Everything together, in one loop.

Lab SDKs see, hear and understand. AHOY decides and acts: routes, alerts and answers, on the edge, on premises or in the cloud.

All SDKs

Research verticals: 5+Core areas of exploration and development.
Experiments: 30+Prototypes and real-world evaluations.
SDKs in production: 3Stable SDKs powering customer workflows.
To cloud, on your terms: EdgeDeploy anywhere: cloud, edge or on premises.

Implementation partners

Trusted by innovative teams

FairPl.ai
wrk
Fly+
Edgecome
Aeromob

SDKs

Five SDKs. Built in the lab, proven in the field.

Each one starts as a research track in the lab and is hardened inside AHOY before it reaches customers. Every live SDK has a playground you can open in the browser.

Compare all SDKs

Point-cloud street scene with vehicles tracked inside detection boxes

Live

Vision

Perception SDK

Cameras in, detections, tracks, counts and events out. Govern the data, gate the model, run it on your own GPUs.

Try it live

A knowledge graph of clustered data points connected by lines of light

Live

Knowledge and reasoning

Data Intelligence SDK

Knowledge graph + RAG, so your applications reason over private data with every answer traceable to its source.

Try it live

Live

Voice

Voice SDK

Every stage of the voice pipeline, self-hosted: clean, transcribe, identify, reason and speak.

Try it live

A grid of wireframe data tokens with a few lit and one raised in orange

Coming soon

Security

Tokenization SDK

Secure data tokenization and orchestration at scale. Currently in development.

A generative point field streaming outward from a dense form

Coming soon

Generative AI

GenAI SDK

Multimodal generative AI capabilities for diverse applications. Currently in development.

Swipe for all 5

Why teams choose AHOY Labs

Research you can deploy. Not a demo you have to rebuild.

Your data stays yours

Every SDK runs on your hardware, on premises or in your cloud. Nothing has to leave the building.

Already in production

Perception, Data Intelligence and Voice run inside AHOY and partner products today.

Built for the edge

Lightweight runtimes made for cameras, sites and devices, not only data centres.

Composable by design

Use one SDK on its own or combine them. Voice already runs on Data Intelligence.

Answers you can trace

Grounded outputs with sources, versions and audit trails your compliance team can follow.

Research

Six disciplines behind every Project.

Research agenda

Agent-based reasoning
Autonomous agents that plan, reason and collaborate to solve multi-step problems.
Graph-based intelligence
Dynamic knowledge graphs that evolve in real time to represent complex world states.
Federated and private learning
Training models across silos without ever exposing the raw training data.
Cryptographic AI
Verifiable computation and zero-knowledge proofs for AI outputs.
Orchestration
Managing thousands of models and agents across distributed infrastructure.
Real-time systems
Sub-millisecond decision making for physical-world control systems.