Contact person: Game Commentary Desktop App

Phone:Show

Email:Show

Location: Sydney, Australia

Budget: Recommended by industry experts

Time to start: As soon as possible

Project description:
"Hello there. I'm going to provide a summary of all the information we've discussed, framed for a software developer.
Project Goal
The goal is to create a single, unified desktop application for Le Mans Ultimate that generates AI-driven, voice-based commentary for race replays. The program will enrich the commentary with historical and real-time data from the SimGrid API, and offer built-in options for uploading to platforms like YouTube or live streaming.
Core Components and Modules
The application will be built as a monolithic desktop application to ensure a seamless user experience, direct access to local files, and optimal performance. It will not rely on a separate server or multiple running programs.
* Telemetry Data Ingestion:
* The program will need a module to read telemetry data directly from Le Mans Ultimate replay files. This is a critical component that will parse data streams for every car, including throttle, brake, steering inputs, speed, and g-forces.
* SimGrid API Integration:
* This module will fetch contextual data via the SimGrid REST API. It will perform GET requests to retrieve information about past and present championships, race results, driver and team standings, and historical rivalries. This data will be used to provide a narrative for the commentary.
* AI Commentary Generation Engine:
* This is the "brain" of the application. It will contain an event detection engine that analyzes the telemetry data stream in real-time. It will use predefined rules to identify key moments, such as overtakes, close battles, driver errors, or pit stops.
* Based on these events, it will formulate natural language prompts that are sent to a Large Language Model (LLM) API (e.g., OpenAI, Claude).
* The LLM's text output is then passed to a Text-to-Speech (TTS) service (e.g., ElevenLabs) to be converted into an audio file.
* Audio and Video Synchronization:
* This is a sophisticated module that handles the timing. It will queue the AI-generated audio clips and play them at the precise timestamp of the corresponding on-screen event in the replay. It will also be responsible for mixing the game's original audio with the new commentary audio into a single, cohesive track.
* Streaming and Uploading Module:
* This module will use a library like FFmpeg to encode the mixed audio and the game's video into a single file.
* It will then use platform-specific APIs (e.g., YouTube Data API, Twitch API) to provide options for direct video upload or live streaming via RTMP. The user will authenticate their account once via OAuth 2.0.
Technical Implementation
The program should be developed using a language and framework suitable for desktop applications that require high performance and direct system access, such as C++ or C# for native development, or Python with a framework like PyQt or Electron for cross-platform compatibility. The core logic should be designed to be modular and scalable, with clean APIs for each component to communicate with the others." (client-provided description)