I’m Siddhanth Balakrishnan, an AI engineer based in Chennai who loves building end-to-end pipelines that turn raw data into usable, high-quality outputs. At OO Studio, I’ve worked across the full workflow—from designing dataset pipelines for video training, to training character LoRAs for repeatable identity control, to evaluating and chaining modern video generation models for longer clips. I also enjoy making these systems practical and reliable in real-time settings. I’ve built production-grade components like lip sync, restoration, upscaling/colorization workflows in ComfyUI, and a real-time multi-speaker dubbing pipeline for live streams (with forced alignment improvements). I ship with strong deployment discipline, using Docker, FastAPI, gRPC, and monitoring with Prometheus/Grafana, while continuously pushing performance and quality in AI video and speech/audio systems.…I’m Siddhanth Balakrishnan, an AI engineer based in Chennai who loves building end-to-end pipelines that turn raw data into usable, high-quality outputs. At OO Studio, I’ve worked across the full workflow—from designing dataset pipelines for video training, to training character LoRAs for repeatable identity control, to evaluating and chaining modern video generation models for longer clips. I also enjoy making these systems practical and reliable in real-time settings. I’ve built production-grade components like lip sync, restoration, upscaling/colorization workflows in ComfyUI, and a real-time multi-speaker dubbing pipeline for live streams (with forced alignment improvements). I ship with strong deployment discipline, using Docker, FastAPI, gRPC, and monitoring with Prometheus/Grafana, while continuously pushing performance and quality in AI video and speech/audio systems.WWWW…