Our passion for experimental user interfaces and our drive to discover new and improved ways of utilizing technology compelled us to envision how a personal voice communicator could function and appear on Apple Vision Pro.

We initiated the project by analyzing the current functionality of voice messaging. The primary applications that support voice messaging features include WhatsApp, Messenger, and Instagram.

However, due to the vastly different nature of the circumstances in which a mobile communicator is used compared to the Apple Vision Pro headset, we had to think completely outside the box.

The headset is typically operated at home and when there is sufficient time available for its use. Starting with a blank slate, we held a single goal in mind: utilizing the spatial interface and a comfortable at-home setting to conceive and develop a genuinely intimate, personal voice communicator that would forge a closer connection between two individuals.

This is when the concept of Closer was conceived—a spatial communicator designed for intimacy, aiming to bring people closer even when they are physically apart.

Our ambition was to employ Artificial Intelligence (AI) to visualize the semantic meaning of the received audio message and immerse the listener in the intended thoughts of the sender during the recording of the note.

We aimed to provide the user with a simple yet intuitive interface to navigate the app. The dashboard view utilizes a side tab bar for switching between the inbox and the recording mode. The UI is streamlined to a minimum, avoiding unnecessary obstruction of the user's view.

Capitalizing on the advantages of incorporating a third spatial dimension, we crafted an intuitive and straightforward recording interface. An audio orb responds to the user's microphone input, providing visual confirmation that the sound is being captured.

Our aspiration was to forge an intensely personal and intimate connection between the sender and the receiver. Here's how we brought this vision to life:

Received messages, while fundamentally recorded audio, undergo analysis by AI and are transcribed into text. Subsequently, the AI generates a spherical image that encapsulates the message's content.

This visual representation is then transformed into a three-dimensional bubble, enabling the recipient to gain an immediate 'sense' of the message's essence.

When a message is played back, we immerse the user in an AI-generated representation of the message's content. For instance, below Agnes is sharing a dream she had, vividly describing her experience of walking through a bustling street in India. The other image shows isImmersion mode for another message—this time from Mike—where he notifies his intention to visit for a dinner with his wife.

While implementing Closer, we dedicated significant attention to micro-animations, enriching the user experience.

We devised a spatial icon in accordance with the visionOS guidelines, causing it to emerge when the user gazes upon it.

The icon's visuals encompass several concepts. The white shapes can be interpreted as the letters 'C', aligning with the app's name, 'Closer.' They are also mirrored, symbolizing two users situated on opposite sides of the app. When combined, the shape at the intersection of the 'C's resembles an eye. Lastly, the 'C' shapes signify the immersion mode, encircling the user in a span of 180 or even 360 degrees.

We developed the application using Xcode Beta, Swift, and SwiftUI.

We designed and animated 3D components, such as the bubble representation of the message, using custom shaders and shader graphs in Reality Composer Pro.

On the backend, we established an auto-scaling infrastructure on the Google Cloud Platform, incorporating a Google Cloud Run service to host our backend API. Whenever a voice message was sent, it was routed through this backend to undergo transcription into text using Google Cloud Speech to Text. Subsequently, we utilized OpenAI's ChatGPT API to analyze the message's semantic content and distill key concepts articulated by the user. We then tasked ChatGPT with transforming these concepts into descriptive scenes. Lastly, we fed these descriptions into Blockade Labs' AI-powered Skybox generator to produce immersive 360-degree scenes.

Get in touch to discuss the possibility of launching your brand on Apple Vision Pro today at
**vision@lepolishbureau.com**

Closer 

Apple Vision Pro  AI-powered voice communicator

Closer

Product

Consumer Electronics

Technology

Projects

About

Capabilities

Process

We are specialising in efficient cross-platform native apps development. Applications that require sophisticated, 3D visual content are developed in [Unity](https://unity.com/), while more traditional UI-heavy apps are done in [React Native](https://reactnative.dev/).

Unity is the leading Game Development platform, but it is also perfect for developing native mobile apps that provide 2D and 3D interactions. We have years long experience in delivering such apps across iOS and Android platforms, while ensuring a single, coherent stream of work. 

Such approach not only optimises cost and time, but also facilitates visual and functional consistency across platforms.

React Native is amazing for another reason. It allows to use the same technologies for native mobile apps development as are used for modern web. Contrary to old approaches that were packaging HTML5 websites into UI-less mobile apps, resulting in poor performance, React Native utilises actual native components of each operating system - iOS and Android respectively.

Regardless of the technology chosen, cross-platform mobile apps development ensures optimised cost, time and reduced risk of platform-specific bugs.

Mobile Apps

Rapid and efficient cross-platform development of the highest quality using React Native, from MVP to the final product.

We are a trusted technology partner for startups, agencies and corporate clients. We build prototype-, MVP- and product-grade solutions in the area of web, mobile, Mixed Reality, Machine Learning and Blockchain.

Mobile Apps | Capabilities | Le Polish Bureau

We are specialising in developing tailored, bespoke Machine Learning solutions for various businesses. In our products, we also incorporate state of the art AI components like Google Cloud Vision or Google Text To Speech.

Our Machine Learning know-how can be utilised in all sorts of domains - from creative-tech to gaming, messaging, human-machine collaborative tools, automatic problem solving, data insights computation as well as data acquisition, augmentation and processing.

Where there's data, there's knowledge. It's just a matter of using the right process to get to it.

Chatbots are a great way of automating communication. They can be text- or voice-based and they can live across pretty much any messaging platform. At Le Polish Bureau, we have experience in delivering solutions for:
- [Slack](https://slack.com/)
- [Facebook Messenger](https://www.messenger.com/)
- [Google Assistant](https://assistant.google.com/)
- [Amazon Alexa](https://www.amazon.com/echo)
- Web

Our natural language processing solutions are based on [Google Dialogflow](https://dialogflow.com/), providing highest level of semantic analysis, combined with support of multiple languages and deployment platforms.

A great example is OnePlus: Crackables, where a chatbot we developed was driving an online puzzle contest.

Sometimes you might have a specific need or an idea that you don't necessarily have all the needed data for. And gathering the data by your team will likely take ages. Probably literally.

This is where our Machine Learning data crowdsourcing solutions come to play.

We implemented this approach in order to analyze people's perception of the phenomenon of the northern lights. We developed a custom interactive questionnaire that allowed respondents to group the auroras by visual similarity. A different questionnaire was then asked to match audio to visuals in order to train the AI about "what an aurora could sound like".

Via data crowdsourcing we obtained over 10,000 data samples within just 24h!

There doesn't always have to be a clear scientific task to be solved with Machine Learning though. AI can be used for creative purposes too.

One of these is Neural Style Transfer. A use of AI to generate completely new visuals or audio based on reference style and a target piece to stylise.

Another interesting domain is reinforcement learning. It relies on the ability to run tens of thousands of parallel, sped-up simulations of a usually mechanical task that the AI is supposed to learn. For example, throwing balls to a basked. Without having to be taught explicitly how to do this, an AI can master such task quickly:

Another domain we specialise in is Machine Vision. We are using state-of-the-art implementations from Google to analyse the world around you. We can also implement custom Artificial Neural Networks that are specifically trained to perform a new detection task.

Regardless of what your challenge is, feel free to reach out. Our Machine Learning know-how is likely to apply.

See how we implemented Machine Learning solutions before:

AI / Machine Learning

Data services, custom models, crowdsourcing, cloud training, vision, language understanding, reinforcement learning.

We build cloud infrastructures for existing and for new projects. This includes RESTful APIs, CMSes, full back end infrastructures, game servers, real-time services, or server-side data processing, video rendering or Machine Learning.

When it comes to scalability and performance, we trust the leaders. All our solutions are based on [Google Cloud](https://cloud.google.com/). Back end infrastructures hosted on Google provide auto-scaling. This means that no matter how many visitors or users you have, the infrastructure will handle that. You don't have to worry about temporary spikes either. What's more, it results in cost optimisation as well, as the platform stops unused resources whenever there is downtime.

Our solutions involve producs like:
- Google App Engine
- Google Compute Engine
- Google Cloud Datastore
- Google Cloud SQL
- Google Cloud Storage

We also have a proven path to delivering state-of-the-art, customisable CMSes for your products developed on top of [Strapi](https://strapi.io/). Strapi is an open source CMS providing a modern architecture. It has recently [received massive funding](https://strapi.io/blog/announcing-strapi-4m-seed-funding-round) and is a product to stay with us for years. All our solutions are completely agnostic of the front end layer of your application. This means that if you ever wish to update or completely re-build the visual side of your project, you will never have to touch the data or the CMS in order to accommodate.

The data stored in our CMSes is available to your application through RESTful and GraphQL APIs.

We also develop custom real-time back ends tailored for games or collaborative online apps. They are based on WebRTC and WebSocket technologies, which allow your users to exchange multiple data packets per second, avoiding typical latency inevitable with HTTP requests.

A great example of that could be the back end powering Start Wars: Lightsaber Escape, which handled millions of players, each using two devices at the same time, where each pair of their devices was exchanging 20 data packets per second.

Cloud Computing

Ultra-scalable data and processing solutions on Google Cloud Platform. Real-time back ends, privacy-focused distributed Machine Learning and data crunching.

We are specialising in professional, modern web development. This includes web apps, product websites, portfolio sites but also highly interactive, animated, 3D pieces of art.

Our team has years long experience working with WebGL, THREE.js and other 3D tools that are necessary to achieve stunning 3D online experiences.

PWAs are highly-optimised web apps that load quickly, support offline access (to an extent at least) and, for example, can be integrated into a mobile phone's home screen similarly to native apps. We develop them in React and we use Service Workers to provide the offline functionality.

We work with WebRTC and WebSocket technologies to implement massively-multiplayer online solutions that respond to user's input immediately.

An example can be the work we've done for Star Wars: Lightsaber Escape, where we allowed users to control a lightsaber with their mobile phones.

We design and develop entire projects, but we are also happy to work on a component basis, whenever your find a challenging aspect of your idea that needs dedicated help.

See an example web project we helped deliver below:

Web Development

Interactive web development using React and the latest tech. Server-side rendering, SEO and performance optimization.

We develop solutions in the area of Augmented Reality, Virtual Reality and Mixed Reality. These include mobile apps using ARKit and ARCore (iOS and Android respectively), as well as PC-based or standalone experiences using AR and VR goggles, like HTC Vive, Oculus Quest or HoloLens. We also specialise in social AR effects, for Facebook, Instagram and Snapchat.

Social platforms like Facebook, Instagram and Snapchat are just the right environment for Augmented Reality. They already have millions of users and they are intended to share and communicate. AR can extend the depth of communication through augmentations that are fun, expressive or even productive.

We also build completely bespoke, native AR apps that can offer much more robust experiences. One of the examples can be the JFK: Moonshot app that we helped develop.

Augmented Reality can also be accessed on the web. We use [Zappar](https://www.zappar.com/) and [8th Wall](https://www.8thwall.com/) technology to deliver marker-based as well as marker-less solutions online, without the need to install or download any app.

See one of our recent AR executions below:

AR / VR / MR

Social AR, Facebook, Instagram, Snapchat, iOS, Android, ARKit, ARCore, MS Hololens.

Capabilities

Mobile Apps

Overview

Unity

React Native

Case Study

Project