Building Apple Vision Pro App with Metatable.ai
Apple Vision has emerged as part of a very crucial stage in the evolution of VR — facilitation and mass adoption of virtual reality technology. Yes, Vision Pro is predated by a bunch of working VR headsets, VR Chat, and even smartphones with some cardboard on them.
But Apple takes things even further by allowing us to actually take an advanced VR headset along and use it outside of our comfy abodes. The main selling point of Apple Vision Pro is its fully mobile nature, allowing for the first-ever convenient, efficient use of VR on the go.
With this novel field of application, we couldn’t help but wonder how our own Metatable.ai’s capacities can power a custom solution created for Apple Vision Pro (AVP).
We decided to run a test case for a hackathon, where we would build an application backend. This resulted in Art Sommelier — an MVP version of a smart app for Vision Pro that generates images based on user input.
Test Overview
Art Sommelier started as an idea to test-drive the Metatable AI platform and see how quickly and efficiently we can build a work-ready backend for an AVP app. This was also the first Apple Vision Pro app in our portfolio ever. Thankfully, the backend launched successfully during the hackathon organized by Lift99.
The main functional goal of the case was to demonstrate the stand-out capacities of both solutions — the device (AVP) and the platform (Metatable):
Apple Vision Pro is the latest, hottest eye-candy tech and the most desired visual tech. That’s because Apple as a whole and this device in particular handle hi-def visuals so well. AVP’s unprecedented image quality AND virtual controls and editing abilities inspired us to create a solution that instantly generates the required images that can be dragged and dropped in virtualized environments (e.g., you can generate a piece of art and drag it over the wall to see how it fits);
Metatable (MT) is an AI-induced powerhouse platform for extra fast (on-demand if you wish) development and deployment of backends. This should serve as an indispensable tool for app creators and designers/developers looking to keep a competitive pace in the market. MT also allows for simple future scaling of the created backend, letting you expand an MVP project without reworking the backend.
Development Process
Working in terms of the hackathon, here are the two main angles of the project we handled to achieve an MVP:
Frontend. While we put the most focus on the app’s backend, we still needed a convenient frontend, so we created a native UI using Swift and connected it to the Metatable backend via two custom APIs;
Backend. The whole backend was thus based on two main APIs for two essential functions to be demonstrated at the hackathon:
The Chat API enables a custom prompt for an OpenAI smart assistant integrated with the MT backend that tunes the assistant to survey the user and recommend pieces of art.
The Image generation API also uses OpenAI’s capabilities to generate images based on the assistant’s descriptions and send them to the AVP app via an MT API.
All this work resulted in a functional MVP with a cool concept — an AI agent for Apple Vision Pro that generates art images based on your taste and preferences. And you can freely fit any generated visual anywhere in your interior using the AR functionality of the AVP headset.
Key Features
Art Sommelier turned out to be quite a feature-rich MVP project, offering:
Personalized art generation based on user preferences: Interconnected backends do the magic of instantly delivering just the art you’re searching for based on even a vague description;
AR visualization in real spaces using Apple Vision Pro: There is an undoubtful convenience in seeing the generated art against the backdrop of your living space (purposes can be plentiful, from virtual workplace organization to classroom studying to simple aesthetics);
Scalability and quick deployment with Metatable.ai: Metatable.ai makes this MVP backend solution easily reusable, simple to maintain, and flexible in scaling, with a clean codebase that can be re-generated and customized autonomously via a simplified, user-friendly toolset.
Combination of AI and Virtual Technology
Our test case is yet another testimonial to the relevance of advanced technology in everyday use — in particular, to the way different types of tech can and should be combined. AI, VR, and AR are all disruptive technologies that help us reinvent the way we deliver and consume services, media, and even aesthetic pleasure.
Pros
For instance, you can combine AI and VR to deliver:
Enhanced user interactions with automated responses and zero downtime;
Personalized shopping or other experiences that match user preferences based on the history of purchases/interactions;
Ultimately efficient data visualization, where smart algorithms easily visualize even large datasets;
Invaluable analytics based on data and interactions tracked in real time;
Tons of automation opportunities that can help specialists get rid of counter-productive routines.
Cons
The only downside of this promising synergy is the seeming complexity and range of effort one must put into actually combining AI and VR to achieve feasible results:
Sophisticated AI operations may require high computational capacities and resource-intensive programming;
Data collection for analysis and personalization must fall under strict data privacy and protection regulations;
Input data for AI models must be high-quality, readable, and accurate for efficient results;
Opportunities
Thankfully, these and other challenges of implementing AI/VR solutions can be overcome all the more easily with emerging platforms like Metatable.ai. Metatable leverages AI itself to help make AI-related processes easier. The processes like backend generation for a novel app.
Thus, using Metatable.ai and a combined AI/VR/AR approach (like an idea to create an app for a VR headset), you can successfully create feasible, useful products for:
Surgical assistance and medical training through VR/AR simulations in healthcare;
Advanced customer experiences (e.g., with virtual showrooms and dressing rooms), navigation, and personalization in retail;
Interactive learning abilities and personalized virtual tutoring in education;
Safer equipment maintenance and real-time process tracking in manufacturing and construction;
Immersive gameplay experiences and virtualized, collaborative content creation in media and gaming;
Virtual tours of properties and visualizations of construction projects in real estate.
Conclusion
Our take on the hackathon participation with the solution we’ve never worked on before has eventually turned into a broad exploration of possibilities that can be unraveled and leveraged through the combination of smart tech and virtual reality. This is a venture we are yet to continue to discover. But the main discovery for now is that Metatable helps to really speed things up and make complex projects simpler to handle. Mostly, through automation and a readymade approach.
We’ll see what more we can squeeze out of the platform. Of course, we’ll share the progress, so stay in tune!