Yuma is developing the most advanced AI agent orchestration platform, dedicated to customer support and ecommerce. We are automating the heavy burden that large Shopify merchants face in their day-to-day activities. This market is vast and homogeneous, and every single merchant needs Yuma. We have a head start in this space, but we will not rest until we are the clear winner in the e-commerce customer support AI market.

Yuma was launched in early 2023 by a third-time YC founder and is already serving hundreds of paying customers.

About Yuma AI:

Yuma is building a complete orchestration platform that deploys autonomous AI agents dedicated to customer support in e-commerce. We are one of the leaders in this space and we offer one of the most advanced platforms to date, supporting over 100 paying customers. Our agents replicate human capabilities by taking actions and retrieving information from external services as needed.

Our top merchants automate over 80% of their support tickets through Yuma.

Yuma was founded by Guillaume Luccisano, a 3rd time YC founder, and now boasts a team of 20 highly dedicated, talented people driven to deliver the best value to our customers through AI.

Role Overview

We remain a small team of 7 engineers, but we are intense and we move very quickly! Fast iteration is key. In the past 2 years since Yuma’s founding, we’ve built a lot, but there is still much more to do! We operate at startup speed and handle everything: product iterations, A/B testing, LLM research, AI Agents pipeline, new feature development, performance optimizations, and more. We are intensely focused on delivering value to our users and on product excellence. That is the only thing that matters.

Our backend is a standard Ruby on Rails API powered by PostgreSQL, Redis, and Sidekiq. Our frontend uses Next.js and React. We already handle a significant load, ingesting and processing millions of support tickets per month and millions of e-commerce data points per day.

We are definitely not a 9-to-5 team, and we are heavily committed to delivering excellence throughout our craft.

The ideal You:

You have led and built multiple large Rails applications, and you always strive for excellence. You are pragmatic, making the right trade-offs that favor speed of iteration while maintaining high product quality. You are driven by delivering the best experience and value to your users. You might have a preference for working on backend projects, but you aren't afraid of frontend work. Rails has no secrets for you.

Recent experience playing with LLMs is now mandatory, and you have already adopted AI in your day-to-day tasks. Building with LLMs and managing non-deterministic output is not scary to you.

If you are relentless, deeply caring about delivering value & ownership, and looking to join an early, hard-working team, we want you 👈. We are seeking someone who is looking to embrace a new challenge and commit to it 100%.

Your diploma doesn't matter to us; we only care about your skills, your sense of ownership, and drive.

The Alternative You:

If you’re still junior, but the description above sounds like the person you aim to become in the next 3 years, don’t be shy: there’s always space for raw talent at Yuma.

OUR CULTURE

Please, if you are considering applying, first read our culture page: https://www.notion.so/yuma-ai/Yuma-s-Culture-5b0e15f1334242ce8a62daab9f2038a1?pvs=4

We have a Ruby on Rails API backend, which is coupled with Postgres, Redis, and Sidekiq. The frontend is built using React and NextJS. Our approach to LLMs is custom-made and tightly integrated with our backend. We have full introspection and logging for every API call we make.

Our approach is heavily product-first, but since we are addressing a problem with new technology, some of our work may be more research-oriented. We don't have all the answers yet on how to solve this problem, and we need to experiment and try out different approaches in a fast, iterative manner.

With already a hundred paying customers, we are already facing some interesting scaling challenges as well.