🚧

UnifyAI is in Alpha

In Alpha, you can expect that UnifyAI is growing and iterating rapidly (even multiple times per day). Some of this documentation may be lagging a bit behind the rapid iterations. We are working hard to keep up with the data science team! If you are having any issues, you can reach us any time at [email protected]

Using UnifyAI - Alpha Playground

UnifyAI is currently in Alpha. To make it easy to explore UnifyAI, we have launched a Playground within the Portal product.

How to find:

You can access Portal through portal.datastreamer.io. If you do not yet have a portal account, account creation is quick and simple.

When you are logged into Portal, you can access the "Playground" under the UnifyAI header in the menu. You can also go directly to portal.datastreamer.io/unify_ai

How to use:

Within the UnifyAI Playground, you can paste any unstructured content. You can then select one of the prepared use cases. These use cases are a selection of prompts created by the feedback our customers have shared with us.

Clicking "Run" The result showcases the response from the UnifyAI prompt.

Example Prompts

• "What is the highest amount of a currency mentioned in the text" - This prompt is an example prompt showing UnifyAI's abilities to read the provided text, scan for numbers or number words, and calculate the highest amount that is present. Similar prompts can be used to generate revenue fields, inventory fields, and other similar numerical metadata fields using the input text content.

Example Data

• "What is the highest amount of a currency mentioned in the text"

this is some sample text with the number of 2 hundred thousand being written in it,
also 100,000 is another number in this text, and one thousand point zero one.

200,000

Example Prompts

• "Return all mentioned dates along with a short description" - This prompt is an example prompt showing UnifyAI's abilities to read the provided text, scan all the dates presented in the text, analyze them, and prepare a short explanation for each date. This brings searchability to unstructured data for use cases when dates matter. For instance, financial documents, grant applications, insurance claim data, legal documents, and so more.

Example Data

• "Return all mentioned dates along with a short description"

These funding programs represent a form of partnership with community based organizations 
which in turn contribute significantly to the goals in relation to community capacity, 
equitable access, well being, diversity, civic participation and civic cohesion.
Events: $7,000 to $25,000; Outreach activities: $7,000 to $50,000; higher amounts can be considered
if well justified. Starting date will be 2023 Januaray first and all the application should be received till 2024/12/20. 

2023/01/01 (starting date)
2024/12/20 (deadline)

Example Prompts

• "Return mentioned foods along with their category based on the food pyramid" - This prompt is an example prompt showing UnifyAI's abilities to read the provided text, scan for specified items, and categorize them based on their characteristics using a reference group. In this use case, it first finds all the mentioned food items and then tries to find an association between each food item and category from the Food Pyramid. Similar prompts can be used for different classification tasks with pre-defined grouped items.

Example Data

• "Return mentioned foods along with their category based on the food pyramid"

This is an ideal thanksgiving dinner, a roasted turkey with stuffing and mashed potatoes
on the side plus hearty salad and plenty of cranberry sauce

Turkey: Protein
Stuffing: Grain
Mashed Potatoes: Carbohydrate
Cranberry Sauce: Fruit

Example Prompts

• "Return publicly traded organizations along with their stock tickers and stock exchange where they are traded" - This prompt is an example prompt showing UnifyAI's abilities to read the provided text and apply a multi-level analysis. First, it scans all mentioned organizations then for the ones that are publicly traded, it provides two levels of information: the associated stock tickers and the stock exchange where the tickers are traded. This ability simply shows the capability of Unify AI in processing a piece of text and using external references to enrich unstructured input text.

Example Data

• "Return publicly traded organizations along with their stock tickers and stock exchange where they are traded"

We view Micrsoft now even in a stronger position from an AI perspective with Altman and Brockman”
at the company, Ives wrote in a note on Monday.

 Microsoft (ticker symbol MSFT), NASDAQ

Example Prompts

• "Return mentioned products along with their types and a relevant description." - This prompt is an example prompt showing UnifyAI's abilities to read the provided input, search for specified items, and generate metadata that can fill a customized schema. In this use case, first, it scans all mentioned products, then it provides types and descriptions as two metadata. This ability simply shows the capability of Unify AI as an assistant tool in enriching unstructured data and improving the corresponding schema.

Example Data

• "Return mentioned products along with their types and a relevant description."

The 13-inch M2 MacBook Air delivers strong performance and fantastic battery life 
at a reasonable price. And the Air’s keyboard is actually good, since Apple traded out 
the shallow and failure-prone butterfly switches under the keys of 2018- and 2019-era MacBook Airs
for springier and more satisfying scissor switches. 
Like Apple’s other laptops, the Air has only a few ports 
(in this case, two Thunderbolt 3/USB-C ports, which may require you to use a hub or new cables). 
But the Air’s light weight, solid construction, and industry-leading support make it 
a great laptop, especially if you also own an iPhone or other Apple devices.

Product Name: MacBook Air
Type: Laptop
Description: Lightweight, solidly constructed, and with industry-leading support, making it a 
             great laptop, especially when paired with other Apple devices like iPhones.