How to get started
Our integrations with QuickBooks Online and Xero are complete! You can now incorporate cutting-edge AI into your business.
You only need a few minutes to connect your data source and start chatting with your data!
Let us know if we can help.
How to register
To create an account with chata.ai is very simple, just follow the step-by-step below.
- Fill out all of the required fields and agree to the terms and conditions. Select “Sign me up as a Partner” if you are an accountant or Bookkeeper and would like to bring chata.ai to your clients. Learn more about the benefits of becoming a chata.ai partner.
2. You will receive an email asking you to complete your registration. If you don’t see this email shortly after registration, please check your junkmail. Once you respond to this email you will be asked to set up your profile and password with chata.ai
How to Connect
Once you have completed your registration, you are now ready to connect your first data source.
- Click “Connect Data Source” from the left-side menu
- Select the organization that the data source belongs to (i.e Client X, or the name of your company)
- Choose which service would you like to connect to (QuickBooks online or Xero)
- Name your data source
- Click “Connect to QuickBooks”
Depending on the size of your data set in QBO or Xero can take up to 15 minutes to connect your data source for the first time.
Attention Innovators: Reconnect your data source
If you connected a data source during our Innovator’s release, you will have to reconnect your data source when logging into chata.ai commercial release for the first time.
Let us know if we can help.
How to add a new organization
Connecting a new organization or client data source allows you to connect to additional data sources in addition to your own personal account. Learn how to create a new organization in this video:
- Click on Connect Data Source from the left menu
- Under the organization drop down, select ‘New Organization’
- Complete the form that asks for the client information. This will allow your client to have access to chata.ai. Your client will receive an email that will inform them that they have been connected to chata.ai
- Follow the same steps as connecting your initial data source
How to add your logo
You can customize your reports by adding your company’s logo.
See the how-to in this brief video
How to invite a new user
To invite a new user, go to the Admin page located on the left side menu.
On the top of the page, click on “Invite User”
Fill out all of the required fields and select the user role in the drop-down menu. For each new user, select the appropriate role based on the descriptions below.
How to edit user permissions
To edit user permissions, go to the Admin page located on the left side menu.
Select the user whose role you would like to update and click on “Edit user permissions” located at the top of the page.
On the drop-down menu of the “edit user permissions” window, select the new role and click on the update button.Please select the new role based on the descriptions below.
How to delete a user
To delete a user, go to the Admin page located on the left side menu.
Select the user you’d like to delete and click on “Delete User” located at the top of the page.
What is a query?
A query is how you communicate with chata.ai using natural “human” language. chata.ai takes this query and automatically translates it to a form that your database will understand and can execute on.
How to create a new query?
Learn how to create a new query in this brief video or you can also read the info below.
To create a new query, type your question into the “New Query” field.
You can then hit Enter to execute the query or select from one of the suggested queries from the intelligent autocomplete by either clicking with your mouse or using the arrow keys to highlight the applicable query suggestion before hitting Enter.
Filtering provides a quick way to find and work with a subgroup of data. You can filter by a categorical label, numerical comparison, or a combination of both.
Filtering by labels
Labels are the specific terms that exist in your data set that are unique but are seen across multiple instances of the data. These can be product names, customer names, vendor names, payment type, countries, etc. You can filter by label by creating a “New Query” using the specific terms you are searching for.
Unsure about how you can create a new query? How to create a new query.
See how to do filtering by label in this brief video or check out few examples below
Filtering by label example: Customer name
Sales for Billy Joe
Sales for PVC pipe
Filtering by multiple labels
You have the ability to query multiple labels in the same query. For example, you can filter by product and city at the same time, or you can filter by customers, country, and payment type.
See how to do filtering by multiple labels in this brief video or see an example below
Filtering by multiple labels example:
Sales for Billy Joe for pvc pipe
Filtering by numerical comparison
Filtering by numerical comparison is when you can use terms such as “greater-than”, “less-than”, “equal-to”, to filter through numerical data.
See how to do filtering by numerical comparison in this brief video or check out few examples below
Examples for filtering by numerical comparison:
Sales greater than $500
Sales less than $200
Filtering in grid format
Grid format is when you have a query result in a format that looks similar to what you would see in spreadsheet software.
When in grid format, you can filter using terms including,“greater-than”, “less-than”, “equal-to”, “and”, “not”, “or”. Please note when using “and”, “not”, “or” that the specific label you use with the filter needs to be spelled exactly like it shows up in the data in grid format.
See how to do filtering by comparison in this brief video or read the how-to below
You can filter by comparison using the filter icon.
1. Scroll your cursor over the top-right corner of the message results. Choose the filtering icon. A text field will appear at the top of each column.
2. In the text field on the top of the column of your choice, use a filtering symbol and/or filtering text to ask your question.
In the table below, please see all the types of filtering you can do using the filter icon.
Types of Graphs/Charts
chata.ai allows the user to visualize their data using graphing/charting functionality. Currently supported types are bar, column, line, scatter, pie, area, heatmap, and bubble chart.
Check out this brief video and learn how to create graphs using chata.ai
When data is returned from a query that chata.ai determines is eligible for graphing/charting, some icons will appear on the bottom left of the results window to give the user the ability to select a type for visualization.
The user has the ability to dynamically move from the original results to any eligible chart types and then back again if desired. Even after a chart is executed from the query area, chata.ai will still provide options for eligible charting types.
Learn how to use dynamic graphing in this video!
A Bar Chart or Bar Plot is normally used to present data where the user wants numerical data aggregated for specific groupings of labels such as in the case where one would want to know the total sales per product. Each product would have its own bar and the length of that bar would be based on the total sales. In bar charts, the x-axis is the numerical data and the y-axis has the label groupings.
Learn how to create a bar plot in this brief video or you can also check few examples below.
Bar plot examples
A Column Chart is the same thing as a Car Chart except the axis are switch which shows the labels on the x-axis and the numerical value along the y-axis.
*Note: both examples below were dynamically changed to column charts as described in the Types of Graphs/Charts section
Stacked Bar/Column Chart
The Stacked Bar/Column Chart is very useful when you want to visualize similar data as in the regular Bar/Column charts, but being able to visualize the components of a quantitative number.
If a user visualizes a query such as “total sales per product”, then a bar/column plot would be great, however if they want to visualize “total sales per product per month” then a stacked bar/column chart would work well for this.
Line Plots are normally used to view changes of a certain variable over time such as total sales per month over the last year. You also have the ability to plot multiple lines as well in the case the user wants to know the same type of sales information but on a per product or per country level. In this case, each line would represent a product or country.
Learn how to create a line plot in this brief video
Example of line plot
Scatter Plot - Coming soon
A Scatter Plot is normally used when you want to see specific data points that involve 2 axis of quantitative data. So if a business owner wishes to see a plot of sales vs. quantity for all product sales, they would see a display of many points with each point representing one transaction.
A Pie Chart is very useful to visualize parts of a whole. An entire pie can be thought of as 100% of something and can enable the user to see all of the separate components involved in the whole such as the proportions of each product out of total sales. Each slice in the pie would refer to one of the products.
See how to do create a pie chart in this brief video
Pie chart examples
Area Chart - Coming soon
An Area Chart is useful in similar parts of a whole scenarios as a pie chart, but the Area Chart does this well when looking at changes over a period of time like in the line plot. If a user wants to visualize the different product components that make up sales over the last year, then an Area Chart would be great for this use. The very top of the Area Chart would be the total sales and then each product component would be in a different color below.
A Heatmap is a very powerful visualization that allows the user to view results of 2 different label groupings combined with a numerical aggregation. Each tile on the heatmap will have a specific colouring according to a scale from a low numerical value in the data to a high one. If a user wanted to visualize total sales per product per payment method or total sales per product per country, then a heat map would be a good choice.
See how to do create a heatmap in this brief video or check out an examples below
A Bubble Chart is very useful for multi-dimensional data just like the Heatmap. Bubble Charts specifically are great when you want to visualize data involving one label grouping and three different pieces of numerical information such as total sales, total profit, and average price per product. The user would see the time frame and customer on the x and y-axis and each bubble would represent the amount of expenses. The diameter or size of each bubble would be dependent on the average expense.
You can count how often instances are repeating in your data set.
See how to count in this brief video or check out few examples below
Here are some example using how many or count.
How many unpaid invoices
Count sales last month
You can also count numbers selecting a range of cells. The results will show up in the lower right-hand side of your screen. If you have any question, check out our dynamic statistics page.
To perform a sum function using chata.ai all you need to do is ask questions using total or sum on a numeric data set and let chata.ai do the math for you.
See how to calculate total in this brief video or check out few examples below:
Sum of sales per product
Total sales per payment method
If you want to quickly get the sum of a range of cells, select the range and look in the lower right-hand side of your screen. If you have any questions, please check out our dynamic stats page.
Returns the average or mean of a set of numerical data points.
To calculate the average just include average or mean in your questions.
See how to calculate average in this brief video or check out few examples below.
Average sales and mean sales per product.
If you ask questions using min or minimum, chata.ai will returns the smallest number in a set of values.
See the how-to in this brief video or check out few examples below.
Min of sales
Minimum sales per product
Returns the largest value in a set of values.
Learn how to calculate max or maximum in this brief video or check out few examples below.
Here are some examples:
Max of sales
Max of sales per product
Variance - Coming soon
Variance is a measure for how spread out a distribution of values are.
To calculate the variance, include var or variance in your new query.
Here are some examples:
Var of sales
Variance of quantity per product
Standard Deviation - Coming Soon
Standard deviation is the square root of variance and is also a measure of how widely values are dispersed across a distribution.
Some example queries are:
Sd of sales
Standard deviation of quantity per product
Summary Statistics - Coming Soon
Summary statistics is a very powerful way to get many different types of statistics on a numerical set of data.
Returns the count, mean, standard deviation, min, 25%, 50% (median), 75%, and max.
Summary statistics can also be performed after filtering for a specific label.
Correlation - Coming Soon
Correlation is the average of the products of deviations for each data point pair in two data sets.
Correlation provide a numerical value from -1 (strongly not correlated) to +1 (strongly correlated). A -1 value would mean that as one variable increases the other will move proportionally in the opposite direction. A value of +1 would mean the two variables move in the same direction proportionally if one is increased the other will as well in the same direction. A value of 0 would mean there is no relation to the movements between the variables.
To calculate correlation using chata.ai you only need to add vs or versus in your queries, for example:
Correlation of sales vs quantity
When more than one cell with numbers are selected, summary stats for those cells are displayed on the bottom-right side of your screen.
See how to use the dynamic stats in this brief video or read few examples below.
For example, if four cells on your data set are selected, and they contain the values 4, 7, 1 and 6, the average, count and sum of these values will be displayed.
While using chata.ai the user has the ability to save results for a specific query using Pins or save the results for an entire session using Reports. The user also has the ability to Comment on a specific results that gets saved with the rest of the Pin or Report.
See how to pin in this brief video or check out the step-by-step below
During a session, the user may only want to save specific query results and not the results of the entire session. This is what Pins are for. The user can Pin their query result, which is Timestamped, and then easily view all Pins at a later time associated with that project.
When the user is on their project and have completed a query and they want to pin the results, this is possible to do by clicking the pin icon on the right-hand side of the query box. It will be white if the query/result has been pinned and will be filled in blue if the query/results have been pinned.
After the user has completed all of the Pins they wish to do, they may review them by clicking on the Pins section in the menu. Please note that the Pins section will be dependent on which project you are in.
Reports are useful when the user wishes to save all of the query results for a particular session for a specific project. These reports will be saved with a timestamp in a reports section for the user to review at a later time.
See the how-to in this brief video or check out the step-by-step below.
Hover mouse over the hamburger menu on the bottom left.
Enter the name of the Report you want to save.
The report is now saved under the reporting session on the left-side menu. You are the only one who can see the reports unless you share it with other users on your organization.
See how to create templates in this brief video
Templates are a great feature for those who perform repeat analysis at different time points. For example, a user wishes to run the same 20 queries at the end of each month, they can save their initial session as a Template. What gets saved for the Template is all of the Natural Language queries that were performed during that initial session. When the next month finishes, the user would just load their saved Template and then execute the queries without any additional typing.
Hover mouse over the hamburger menu on the bottom left and click on Save Template
Name template for future use and then select create.
Templates can be found in the main menu section under Templates.