We have worked with AI technologies from the beginning of the new AI era back in 2017, and developed a solution to radically optimize any coding/development/programming process.
Massive total cost of ownership (TCO) savings
Twice the deliverables per one tech talent
Applicability to any project (new development, liveops, legacy)
Significantly faster time-to-market
High level of development process automation
Stanford University's Human-Centered AI initiative presents compelling data on the usage of coding copilot AI solutions by the technical teams.
CODE MAESTRO isn't just a tool. It's your development project's best ally, offering a two-fold functionality that transforms the way you approach coding and documentation:
🟡 Creates new big parts of code fully automatically (such as entire new classes)
🟡 Grants rapid understanding and helps develop the legacy code further
🟡 Provides insights on the various aspects of the project mechanics, optimizes code and fixes bugs
🟡 Transforms any codebase, no matter the complexity, into an interactive, ever-evolving knowledge base.
🟡 Allows to dive deep into the project's architecture, making every piece of information readily accessible and navigable, thereby simplifying the understanding and development process.
All data is stored locally on the client's side, no data leaks, maximum privacy | Yes | No | No |
Generation or modification of the code with the possibility of using existing functionality and preserving project's architecture | Yes | No | No |
Focus on game development specifics | Yes | No | No |
Integration with a project's environment (engines, plugins, settings) | Yes | Partially | No |
Communication and support throughout the project. From descriptions of implemented mechanics to instructions for implementing a particular plugin into the current project with an example from existing scripts | Yes | No | Partially |
Assistance in working with a specific script in a dialog mode | Yes | Yes | Yes |
Assistance in working with a specific script in a dialog mode, taking into account the context of the project and recommendations for the code | Yes | Partially | No |
The possibility of adding new tools (great prospects for expanding and customizing the toolkit), including custom tools | Yes | No | No |
Review of the pull requests | In progress | Yes | Yes |
An internet search tool | Yes | Yes | Yes |
CODE MAESTRO isn't just a tool. It's your development project's best ally, offering a two-fold functionality that transforms the way you approach coding and documentation:
🟡 Code Generation Tailored to Your Project: Leveraging your project's codebase, CODE MAESTRO generates new code that seamlessly matches your project's unique coding style and most importantly fits your project’s architecture.
🟡 Intelligent Script Generation: It can produce scripts that utilize functions from various project parts, facilitating modular development and innovation with AI's help.
🟡 Expanded Contextual Understanding: With an advanced model capable of handling complex and lengthy code, CODE MAESTRO addresses intricate queries with precision.
🟡 Plugin Integration: Recognizing and integrating necessary plugins automatically, CODE MAESTRO ensures your project's environment is always optimized and complete.
code transformed into the knowledge base and documentation generated
docs integrated into database to be used by a ChatGPT style neural network
neural network creates a conversational interface
PROFIT!
🟡 Interact with documentation
🟡 Get assistance in writing code
🟡 Get insights on the various aspects of your project’s mechanics
🟨 PHASE 1
Understanding your project's scope, objectives, and requirements.
Comprehensive study of:
Project Architecture
Scripts
Existing Documentation
🟨 PHASE 2
Knowledge Base Creation:
From the parsed data, our tool automatically generates comprehensive AI documentation. This documentation includes detailed descriptions of functions, classes, methods, and their interactions within your project.
🟨 PHASE 3
Neural Network Training:
We integrate the structured documentation into a neural network model, similar to ChatGPT. This model is trained specifically on your project's data, ensuring it understands the intricacies and nuances of your codebase.
🟨 PHASE 4
Adaptive Learning:
The tool is designed to learn continuously from the new inputs, updates, and user interactions, ensuring that the documentation and assistance it provides are always up-to-date.
🟩 PHASE 5
Final review and usage
Riftwardens is an innovative open-market crypto-collectible game that combines a unique experience of classic RPG with the immersive world of NFTs. The game is in the active development, backed by Social Discovery Ventures, Hitech Ventures, Blackshield Capital and angel investors.
After gaining the first 30.000 people in its gaming community and raising the first round, the team has faced a challenge: rapidly achieve a soft launch product quality to provide the community with the beta release it was looking forward to.
To address the production challenge and meet the deadline, Riftwardens team is using Code Maestro in 2 areas:
🟡 Shader generation
🟡 Assistance in setting up and integrating animations
🟡 VFX pool generation and set up
🟡 Fighting scenes development
🟡 Assistance in building a resting area system for the characters
🟡 Generating of the code with functions and classes that are already present in the codebase
Alex, Technical Artist
Riftwardens
The only thing Alex has to do is request Code Maestro for Assistance in the dialogue window.
Step 1. Alex is asking the Code Maestro to provide a tutorial for creating a VFX pool.
Step 2. After getting a tutorial, Alex requests a code to be generated.
Step 3. To complete the full cycle of the assignment, Alex requests a tutorial on how to integrate the generated code into his project.
One day of the developers’ time saved each time a graphic component needs to be integrated
One or more days of his time are saved when generating or optimizing custom shaders
At the end of the month, Alex completed 2x+ of the planned tasks and became a top performer on the team.
Antony, Unity Engineer
Riftwardens
With CODE MAESTRO, Antony onboarded to the new system in no time
🟡 CODE MAESTRO provided guidance on the existing code base, its syntax and functions. This has shortened Antony’s onboarding from roughly 1 month to 3 days.
🟡 CODE MAESTRO explained the project mechanics and helped generate the classes using the context of the existing codebase.
🟡 Even though the system is new, it is fully adjusted to the project’s architecture, code style and integrated into the project.
Antony is shipping twice more deliverables than his PM initially planned.