Aniket Tayade

I'm a Senior AI Engineer based in Pune, India, working at the applied end of generative and agentic AI on Google Cloud.

I started out in computer vision. At Tarsyer I spent nearly three years building deep learning systems that ran in the real world on edge devices, not in notebooks: warehouse management for Mahindra Logistics that improved turnaround times by 70%, retail analytics deployed across 3,000+ Tata Trent and Bata stores, and safety and inventory monitoring for JSW Steel and Birla Cement, including a conveyor bag counter that hit 99.97% accuracy through motion blur and dust. Working with TensorFlow, TFLite, and PyTorch under real hardware constraints taught me a discipline I still rely on: a model only matters if it survives production.

I now lead agentic AI delivery at Cloud Ace, a Google Cloud Premier Partner, where I drive the India division's practice and lead client projects for our Singapore team alongside Cloud Ace Japan and Google product teams. My day to day is building production Gemini Enterprise agents with Vertex AI Agent Builder and ADK, designing RAG architectures, multi-agent orchestration, and custom MCP tooling. Recent work includes an AI agent suite for a Class III medical-device manufacturer operating under FDA and PMDA regulation, an identity-aware operations assistant unifying POS and knowledge sources for 50+ users, and an agentic ticket-demand forecasting system for a cinema chain.

I hold Google Cloud Professional certifications in Machine Learning Engineering and Cloud Architecture. I was lead engineer on the platform that won ReTech Startup of the Year 2024, core developer on a solution recognised in the Mahindra Logistics Catapult S4 accelerator, and contributed to Cloud Ace being awarded Google's "GCP Champion of the Year" at the JAPAC Partner Forum 2025.

What genuinely interests me is the engineering around LLMs rather than the models themselves: retrieval quality, evaluation loops, guardrails, and the architecture that makes an agent reliable enough to trust in a regulated environment. I'm looking for a remote role with a globally distributed team where I can own AI delivery end to end.