🙅‍♂️

It's a custom sega game. youtu.be/YF9TeUDyXKE?t=415

Yes we do. opensourcepledge.com/ blog.sentry.io/another-year-… blog.sentry.io/we-just-gave-… blog.sentry.io/join-the-pled…

What open source projects should we include?

@waybackmachine the wayback machine page is firing thousands of requests (eventually getting 429) when I scroll the calendar into view for a URL.

I made a Syntax themed Game Boy Color game. youtube.com/watch?v=POylAyrY…

Check out @omidotme fully open source and they have a dev kit omi.me/products/omi-glass-de…

Code for jev chat is available here: github.com/w3cj/jev-chat The README includes a breakdown of how this all works and includes an FAQ.

Requires an extra request to resolve spelling errors, but still works. Combined with cspell and compromise nlp pre-processing, jev can decide what the user actually meant.

The answer is directly from the article right now. It detects a wikipedia lookup. Gets the article. Then does a second request that includes the full article where each sentence is an option for jev to choose. See their guide here: docs.typesafe.ai/cookbooks/s…

I'd have to create a new "timer" tool and make it available to jev but definitely doable! Jev would just respond with which action to time delay, and how long. Then we pass that off to a background queue that invokes it when necessary. Compound requests are a bit tricky though. The smart home example from typesafe AI actually detects a compound request, passes to an llm to de-compose into individual commands. Then passes each command to jev.

Nope! That's just a wikipedia lookup. Jev extracts the topic the user is asking about, queries wikipedia, then a follow up is sent to jev with the content of the wikipedia article asking which line from the article answers the user's prompt.

I pass a "choice" to jev - "Which tool would fulfil the user's latest request (taking the conversation into account)?" and the options are a dynamic list of every available tool. For arg extraction / matching on the user prompt there is a seperate question for every single available tool like: "For a weather request: which city or place does the user want the weather for? If they don't name one, pick the place from the earlier request." and the options provided are extracted from the user prompt as a list of options. Could be the prompt split on spaces, could be smarter word pairs.

I provide an option to jev for home assistant "no device or room is mentioned at all" - this then asks for a follow up that extracts and matches the name to a device in the home. We can do a similar thing for ambiguous tool calls. Provide jev an option "More than one tool call can satisfy this request" and then give a follow up prompt to choose, OR detect high match percent in multiple tools and automatically give a follow up to the user before executing.

Watch the full explainer: youtube.com/watch?v=QbYBRjOa…

I built a chat bot with jev, no LLM at all! Responses are instant, no hallucinations. I hooked it up to web search, wikipedia, weather, todoist and home assistant. Jev decides what tool to call and what args to use based on the prompt. Instant answers cite sources as well!

Jev is an extremely cheap and fast classifier that can replace long LLM prompts in your workflows with more deterministic answers and actions. I break down what that means, how it works and show several demos of Jev in action in my latest video. Watch: youtube.com/watch?v=QbYBRjOa…