Software Engineer: Voice + LLMs

Several years ago Lundy Inc. developed a real estate search by voice application, using traditional NLP. This was successful in the market with paying customers and several million real estate listings syndicated to voice.

By focusing on our core mission of helping the blind and visually impaired, we ended up building a product that was kind of cool for everyone. But it was a bit rigid as a user experience. You had to say certain keywords or phrases for it to work. 18 months ago we began integrating LLMs into the app, and 12 months ago we released the first LLM-based version into production. At that point the LLMs were just an add-on while the majority of the app was still traditional phrase-based NLP. Still, it let us see the potential so we embarked on a rebuild to fully leverage LLMs. This launched on April 11 as Finding Homes v2.0. (Try it out on your Alexa device: "Alexa, start Finding Homes" or see the launch event video https://youtu.be/NavBqhyL4TY?feature=shared). This was a big leap for the industry and got us a lot of attention, and requests from bigger companies to build on and with our technology.

We've really been enjoying blazing ahead with a tiny team of a few engineers. But it's time to hire. Our stack is text to speech and speech to text, nodejs on the frontend, python and postgresql on the backend, and a lot of LLM integration. Our infrastructure is AWS managed with Terraform. The existing products are centered on the Alexa platform but newer products are branching out. The work with LLMs currently is mainly finding optimal, performant, and sophisticated ways to use APIs provided by OpenAI, Anthropic and others, and less about building and running our own, although as we scale up that becomes more attractive.

Experience with the above technologies is a good thing of course. But really things are moving so quickly that the most important criteria are more general engineering mindset and experience, plus the ability and desire to embrace change and learn continuously. What we have to offer is: a little bit of a head start in working with these technologies, a fun and supportive work environment, and the experience of moving beyond learning and prototyping and into production release of impactful LLM-based products.

Apologies for the broad salary range, it's because we are open to a wide range of experience levels for this hire.

The team is very much remote and international, but we would also like to build up the ability to work together face to face where it's helpful, so applications from the home town of our CTO (Charlottesville Virginia) or CEO (San Diego) will get special attention. I think the right person for this job is someone who has a good professional software engineering skills, has been partially working with LLMs, and is dreaming of diving into that world full time. That was my experience, just shifted back in time a bit. I had gotten a bit jaded, but I can honestly say now I'm super excited about work again, can't wait to start every morning. I hope we can do the same for you.