How to Use Jev for AI Decision Making

TL;DR

Jev is a new AI classifier that processes inputs and outputs probabilities for each choice. It operates extremely fast, categorizing 1,700 emails for just 18 cents. Jev excels in making quick, repeatable decisions on incoming data, suitable for business applications like lead scoring and support routing.

Transcript

Jev is here and it's a big deal. It was created by Dooo Almeida. Yes, that's the same guy whose research built chatbt. Now, it's such a big deal because it's a whole new way to do AI. So, I brought on my friend Ryan who's on the founding team of Open Code to just come on and clearly explain what Jev is, what are some insane use cases, and break dow... Read More

Key Insights

• Jev is a classifier that returns probabilities for each choice based on defined input and output schemas.
• A demo showed Jev categorizing 1,700 emails for 18 cents, highlighting its cost-effectiveness.
• Each Jev query takes around 200 milliseconds, making it ideal for fast decision-making tasks.
• Jev is best used in roles where businesses make quick, repeatable decisions on incoming data.
• For high-intelligence tasks like trading, frontier models should be used instead of Jev.
• Instant access to Jev is available through the Vercel Gateway, while direct access requires a waitlist.
• Jev can transform contact forms into decision-making tools, scoring leads from 0 to 1 for quick responses.
• Jev's structured output is type-safe, allowing developers to integrate it directly into code.

Questions & Answers

Q: How does Jev work as an AI classifier?

Jev functions as an AI classifier by taking a defined input and an output schema, then returning a probability for each choice. For example, given an input like an email, Jev can categorize it into various types and assign a probability score for each, allowing for quick and efficient decision-making in business processes.

Q: What are Jev's main advantages in business applications?

Jev's main advantages in business applications include its speed, cost-effectiveness, and ability to make quick, repeatable decisions. It processes each query in about 200 milliseconds and can categorize large volumes of data at a low cost, making it suitable for tasks like lead scoring, support routing, and email categorization.

Q: How is Jev different from traditional chat models?

Jev differs from traditional chat models in that it does not generate text or provide reasoning. Instead, it delivers structured output based on probabilities, making it ideal for decision-making tasks rather than conversational AI. This allows developers to integrate Jev's output directly into their code without additional processing.

Q: What are some practical use cases for Jev?

Practical use cases for Jev include lead scoring, where it evaluates the quality of potential clients; support routing, where it directs inquiries to the appropriate teams; and email categorization, where it sorts emails by priority, category, and spam likelihood. These applications benefit from Jev's fast processing and low cost.

Q: How can businesses get started with Jev?

Businesses can get started with Jev by accessing it through the Vercel Gateway, which provides instant access. For direct access, a waitlist is available. Businesses can experiment with Jev by integrating it into daily workflows that require quick decision-making, taking advantage of its speed and cost-effectiveness.

Q: What limitations does Jev have?

Jev's limitations include its unsuitability for high-intelligence tasks like financial trading, where more advanced models are required. It is best used for routing-style decisions where speed and efficiency are prioritized over complex reasoning or text generation, making it less effective for tasks requiring deep analysis.

Q: How does Jev handle email categorization?

Jev handles email categorization by processing the entire email object, including subject, body, and sender information, and categorizing it based on predefined options such as work, marketing, or spam. It assigns a priority level and spam score, enabling efficient sorting and management of large volumes of emails.

Q: What startup opportunities does Jev create?

Jev creates startup opportunities by enabling businesses to automate decision-making processes at the front of their information queues. For example, a local services platform could use Jev to match customer requests with service providers instantly, transforming 'instant quote' forms into truly instant responses, enhancing customer experience.

Summary & Key Takeaways

• Jev is a classifier AI that processes inputs and outputs probabilities for each choice. It is fast and cost-effective, as demonstrated by categorizing 1,700 emails for 18 cents. Jev excels in making quick, repeatable decisions on incoming data, ideal for business applications like lead scoring and support routing.

• Unlike chat models, Jev returns structured output without visible reasoning, making it suitable for roles where fast decision-making is crucial. Its speed and low cost allow for widespread experimentation, with a $5 intro credit lasting for two days of heavy use.

• Jev can be integrated into various business processes, acting as a traffic cop for incoming information. It quickly determines the importance of data and routes it accordingly, making it a valuable tool for businesses needing efficient decision-making capabilities.