Create data visualizations without ever needing to leave Slack or Hipchat


Create data visualizations without ever needing to leave Slack or Hipchat

Chatlytics analytics assistant, Fetch, works along side your team in Slack or Hipchat, always ready to help sweat the small stuff whenever you need data visualized. Simply ask Fetch to pull up your visualized data right into your Hipchat or Slack chat client.

Chatlytics is like having a BI tool and an analyst packaged together, and seamlessly integrated into Slack or Hipchat workflows. It is simple, flexible and can be customized to fit the style of any team. This allows teams to get more value from the data they already have.

Avatar?ixlib=rails 2.1
I had a look the screencast and I wonder if having access through Slack outweighs the disadvantage of having to learn a new syntax? Wouldn't it be easier to create data visualizations through a more visual UI and then share them on Slack? In my experience designing a useful data visualization is a matter of trial and error which isn't something you would want to bother your colleagues with.

Curious to hear what your thoughts are on this?
Avatar?ixlib=rails 2.1

Great question.

As you alluded there are typically those who create data visualizations and those who consume them. Those who create are those who give context to data. Sometimes creators and consumers are the same people, many times they are not.

This creates a interesting dynamic around workflow. Here is a typical example:

A coworker requests something from a team member to visualize some data. That team member is usually someone with the tools and knowledge to create a data visualization. Lets say they use a tool like Tableau. Once they complete a visualization it is exported as JPEG, PNG, PDF or similar. They send it via email or paste into Slack (or both). In some cases, they may also publish it to a dashboard somewhere (i.e. Tableau Server). Assuming the creator understood the request, the visualization might be exactly what the coworker was looking for. Maybe more revisions are needed.

This is a "synchronous" model, where outputs are "pushed" from a creator to a consumer. In some, not all cases, the assets are static (JPEG, PNG, PDF...). However, they may be able to push a "live" visualization somewhere as part of the process. While this "live: visualization solves the iterative nature of getting updates, it means the person requesting the visualization has to switch contexts (From Slack to Tableau Server). This creates friction in the end users workflow (e.g, leave slack, login to this dashboard, find what they need, create a export and then share back to Slack....). This is where we think Slack, or similar tools, can reduce fiction for both creators and consumers of data.

Rather than focus on static assets or assets published to a server somewhere, a creator can load visualization definitions into slack. Those definitions are there for everyone to consume, when they want to consume them. They also have flexibility to adjust styles, types and variables in creating a visualization. A creator provides the recipe in the form of a query, not a final artifact for each of those variations. This allows an end user the flexibility to create a visualization on-demand, based on those definitions, with live data. They can do so without the friction of being force to switch to a separate context to see a visualization or continually ask someone to send over a new visualization based on variations in there request.

In regards to the syntax, every bot has some syntax, either explicit or implicit. This is something we expect to be refined over time. Assuming users want more of NLS approach, that can be incorporated. However, there are cases where NLS can get in the way of tasks. There is a balance, we think, that needs to be struck. Our goal, for creators and consumers, is coming up with an easily understood, consistent and repeatable method to accomplish a task.

Would love to hear more about your experiences. Any complexities you have faces or observed. Any challenges as teams had grown in size and in different geographic locations.


Avatar?ixlib=rails 2.1
If you need help in SW development, contact me at a.novolaev@elinext.com We have great opportunities for you! www.elinext.com

Sign in with Twitter to join the discussion

Discover startups similar to Chatlytics