Software Engineer at Galleri5. Joined when we were just a few people, so I've worn many hats - backend architecture, ML research, DevOps, you name it. Core work revolves around backend systems and AI/ML - everything from classical ML (NLP, clustering) to cutting-edge generative AI for video and image creation.
Built the backend and AI pipelines that power actual production content. The work is currently being used to create shows streaming on platforms like JioCinema/Hotstar, with more projects in the pipeline. This involves designing autonomous AI agents that can interpret prompts and coordinate between different models, plus building the internal tooling to make these workflows actually usable by the team.
On the R&D side - working with ComfyUI workflows, implementing LoRAs, finetuning models, and building new generation pipelines. Basically handling the full stack of AI content generation - from experimenting with new techniques to making sure everything runs smoothly in production.
