Vishwas Tripathi — Software Engineer

Experience: * For google search infrastructure I have designed microservices framework where needed, and turned down where it was not, overall simplifying backend architecture to stream line production environment and feature development experience. * My current side project include experiments with building RAG infrastructure with multiple input streams (files, search, humans, etc.) * I have worked in google cloud enterprise in redesigning inconsistent (for long time) and extremely slow piece of policy storage and fanout engine, answer was use Queues. Project winning Google engineering of the year. Skills: * Designing big complicated looking systems on simple architecture. Intuitive understanding of things like: * Cache: At all levels from CPU/GPU to Database to RPC * Queues: When to avoid batch jobs, and give better developer experience. * Microservices: Whent to and when not to. * Good grasp over traditional algorithm and datastrucures (Candidate master on codefoces in 2016) * LLMs and neural networks understanding in general, have understanding of design, and optimisation algorithm: * Limitation of tokenizer, cost of higher order optimisers, Kernel search, JIT in GPU, Transformer VS RNN, ... * Limited understanding of frontend design choices and framework, server side, client side, partial reloads, etc. Current interest: Fascinated by LLMs and have understanding of what works and what does not work. Currently exploring to improve this understanding by application.

Stackforce AI infers this person is a Backend Infrastructure Engineer with expertise in cloud solutions and microservices architecture.

Location: Bengaluru, Karnataka, India

Experience: 9 yrs 9 mos