Some of our AI inference are models that we host on SageMaker. Things in AWS can be tricky to configure, but overall I'm amazed how easy they make it to run and scale models.
We absolutely could not build this product with LangSmith for traces, datasets, annotation queues, and evals. We've become big power users! I use to build AI apps before nice tracing tools like this existed and it's like night and day having a tool like this. We considered Langfuse, but I already had experience with LangSmith so for speed purposes we went with LangSmith.
Most of our AI inference is built on Claude (though we use a range of models from OpenAI, Gemini, even Groq). Major shout-out to Anthropic for their models being so exceptional at the tasks we've been giving them. For the most part they can do zero shot on our tasks.
Would love to give a major shout out to our scraping partners, Reworkd. They launched recently on Product Hunt. They've been excellent every step as we've built. We considered using bigger scraping companies like Zyte, but we didn't want to use a large company. I get to talk to the Reworkd founders every day in Slack.
Astrid: Personal Shopping Agent
