Anuj here. My co-founder Amir (Aazo11) and I are building Dataherald (https://dataherald.com/). After our initial HN launch (https://news.ycombinator.com/item?id=26194440) last year, Dataherald is now used by ~100 news outlets nationwide. We are now opening up the product to anyone who wants to tell a story or conduct analysis that leverages data, with a Free Trial (https://dataherald.com/get-started) for the HN community.
Our software makes it easy for anyone to find cleaned, ready-to-use data, and then create auto-updating visualizations from it. All in under 60 seconds. No code required. We’re similar to your favorite data visualization tool (Tableau, Datawrapper, etc) if it already came with the data attached.
I’m a non-technical founder, without a CS background. Over the last several years whenever I’ve wanted to use data to tell a story or conduct analysis that leverages data, I’ve had to rely on a data engineer to create data pipelines to my most heavily used data sources (in my case, Bureau of Labor Statistics). I always think that the data is going to be easy to access and update, but it never is. I’m constantly spending all my time getting the data in a usable format rather than actually doing any analysis. For a public dataset like BLS, haven't a bunch of people already done this? Why can't I just search and find it in normal human language? As an example, I was doing research on the potential employment impact of autonomous trucks on truck drivers (number of jobs lost etc.) and gave up while trying to decipher which BLS series ID I should be pulling.
I’m not the only one. Journalists, for example, feel this pain whenever covering data-heavy topics like the coronavirus, the economy, and especially with data sources that update frequently like real estate prices. The problem is getting larger. There’s an increasing amount of data in the world but most people can't use it to tell a story or make a decision. They lack the ability to build and maintain data pipelines.
With Dataherald (https://dataherald.com/), we’ve developed our own data pipeline engine that can be easily configured to extract data from online sources continually feeding data from public sources and private partners into our data warehouses. We create configurable templates (called Datafeeds) which users can select a few parameters for, and then have an auto-updating visualization. For example "Median home price change" (https://embed.dataherald.com/v3/viz/61b1132960a58f47321bcbc7) is a Datafeed, and the user can select the city and state. We are creating the Datafeeds and setting up the pipelines ourselves for now, but will be opening this up and looking for beta users.
Dataherald streamlines data workflows, establishing a real-time data pipeline connection for users to the data sources they need to run their businesses or tell a story. Non-technical users are more autonomous and are able to diminish their dependency on other teams.
Whether you are a marketer, a researcher, or a journalist, Dataherald helps you tell stories with data faster. We’ve created a Free Trial (https://dataherald.com/get-started) for the HN community, valid for sign ups before Christmas (Dec 25).
One question that we always get that I will clarify now is that Dataherald specializes in helping you harness the power of data that is EXTERNAL to your organization, as opposed to your internal operational data. We’d love your feedback and suggestions.
A small observation here. I don't subscribe ever to any Saas that doesn't makes clear its pricing upfront. 'Start for free' is not exactly clear for me. Also, you might want to put up at least an orientation list, what one would expect on the free data feeds and on the paid ones. Just to help your potential users to make an educated decision. Other than that, I think you have a good business case.
Interesting. If the problem is just finding the data, why not make the product a way to load data into "your favorite data visualization tool". I'm not really the target customer, but will people really want to use yet another data viz tool (idk).
Amir here. We allow users to download the cleaned data and load it into any visualization tool they want. That being said, we have found a lot of users (especially those who tell stories with data) have a lot of use for having the visualization as part of the same solution.
few other things that users typically want that standard data viz tools don't make easy -- auto-updating data pipelines so that your data viz is never stale, and notifications when the data updates. Not exhaustive by any means, but just some more color.
Our software makes it easy for anyone to find cleaned, ready-to-use data, and then create auto-updating visualizations from it. All in under 60 seconds. No code required. We’re similar to your favorite data visualization tool (Tableau, Datawrapper, etc) if it already came with the data attached.
I’m a non-technical founder, without a CS background. Over the last several years whenever I’ve wanted to use data to tell a story or conduct analysis that leverages data, I’ve had to rely on a data engineer to create data pipelines to my most heavily used data sources (in my case, Bureau of Labor Statistics). I always think that the data is going to be easy to access and update, but it never is. I’m constantly spending all my time getting the data in a usable format rather than actually doing any analysis. For a public dataset like BLS, haven't a bunch of people already done this? Why can't I just search and find it in normal human language? As an example, I was doing research on the potential employment impact of autonomous trucks on truck drivers (number of jobs lost etc.) and gave up while trying to decipher which BLS series ID I should be pulling.
I’m not the only one. Journalists, for example, feel this pain whenever covering data-heavy topics like the coronavirus, the economy, and especially with data sources that update frequently like real estate prices. The problem is getting larger. There’s an increasing amount of data in the world but most people can't use it to tell a story or make a decision. They lack the ability to build and maintain data pipelines.
With Dataherald (https://dataherald.com/), we’ve developed our own data pipeline engine that can be easily configured to extract data from online sources continually feeding data from public sources and private partners into our data warehouses. We create configurable templates (called Datafeeds) which users can select a few parameters for, and then have an auto-updating visualization. For example "Median home price change" (https://embed.dataherald.com/v3/viz/61b1132960a58f47321bcbc7) is a Datafeed, and the user can select the city and state. We are creating the Datafeeds and setting up the pipelines ourselves for now, but will be opening this up and looking for beta users.
Dataherald streamlines data workflows, establishing a real-time data pipeline connection for users to the data sources they need to run their businesses or tell a story. Non-technical users are more autonomous and are able to diminish their dependency on other teams.
Whether you are a marketer, a researcher, or a journalist, Dataherald helps you tell stories with data faster. We’ve created a Free Trial (https://dataherald.com/get-started) for the HN community, valid for sign ups before Christmas (Dec 25).
One question that we always get that I will clarify now is that Dataherald specializes in helping you harness the power of data that is EXTERNAL to your organization, as opposed to your internal operational data. We’d love your feedback and suggestions.
Best, Anuj