ARTIFICIAL

ARTIFICIAL

ARTIFICIAL

REALITY

REALITY

REALITY

IS HERE.

IS HERE.

IS HERE.

IMMERSIVE

VISUAL

INTERACTIVE

PERSONALIZED

We are TWO.

We are engineers, designers and innovators who come from MIT, Harvard, Oxford, IIT Bombay, NASA, Google, Samsung and Microsoft. After text and voice, we believe the next chapter of AI is visual, lifelike and personalized. We are building Immersive and magical consumer experiences powered by our AI Engines.

Magic of AI.
In two places.

We trained and developed proprietary machine-learning based software engines to power the next generation of magical and immersive AI experiences.

GENIE

Generative Intelligence Engine

HALO-2

Reality Synthesis Engine

SAY HELLO

SAY HELLO

SAY HELLO

GENIE

GENIE

GENIE

GENIE

Generative Intelligence Engine

Leveraging the raw power of pre-trained Language Models, GENIE enables memory, persona, and personalization.

Memory

Forget everything you know about memory. GENIE is built with dynamic and retrievable long-term memory, maintaining the context of ongoing conversations while recalling past discussion topics to enrich future conversations.

Rich Persona

One of a kind AI. Unique personality. Each AI powered by GENIE has a unique backstory and personality with different preferences that influence every response. Personas evolve over time as they grow deeper with every conversational experience.

Personalized

AI that knows you. AI that you know. GENIE enables AI to understand users as individuals by learning about them, enhancing personalization in chat. GENIE can have millions of different context and memory, unique to each individual users and session.

Step into the world of human like conversations, with memory, persona, and personalizations. GENIE(Beta) is available as REST API or standalone self-hosted license for testing. General availability of GENIE is coming mid-2023. Contact us for demo, early-access and pricing.

Ready to see GENIE in action?

& NOW

& NOW

& NOW

HALO-2

HALO-2

HALO-2

HALO-2

Reality Synthesis Engine

With the power of TWO's ML techniques, HALO-2 generates PixelReal visuals and voice in real-time. 

PixelReal. HiFi.

Meticulously detailed. Uncannily realistic. HALO-2 renders the visuals and emotions of a photorealistic digital character in 4k. It animates the gestures and movements in-sync with natural sounding synthesized audio.

Looks, and smiles.

A variety of looks. All kinds of expressions. HALO-2 enables diverse customization through wide-ranging looks, expressions, and emotions to generate natural interaction. It renders content from multiple viewpoints to support various scenarios.

Realtime. Million-scale.

Interactive. Instantaneous. Wherever. Whenever. Like reality, there is no delay in response when you speak or interact with HALO-2. Users can experience real-time interactivity through high-fidelity realism in ultra low latency.

HALO-2 is built on our proprietary generative AI models and neural rendering techniques, developed at TWO.


Interested in building interactive photorealistic environments or lifelike immersive games? API and SDK access to HALO-2 is planned for the second half of 2023.

ANNOUNCING…

ANNOUNCING…

ANNOUNCING…

Apps &
experiences
of twomorrow.

Our AI engines - GENIE and HALO-2 - are already powering immersive and interactive experiences. We are bringing the power of our tech to TWO's own Apps and Services.

AI messaging and social

app for Gen-Z

Generative AI powered
personalized video messages.

Search. Personalized.

AI powered encrypted personal search.

[ Drag Me Around ]

APP

APP

APP