Crushing the “Time To Analytics” With Data Apps

Implementing Data into your Organization:

“The Cycle of Pain”

1. The business user requests information or a report from a BI developer
2. The BI developer checks to see if they have the data, if they don’t, they go to the DBA.
3. The DBA tracks down the requested data, checks security protocols, and perhaps makes the data available to the BI developer
4. The BI developer creates the report or possibly requests the correct data (again).
5. The business user consumes the report or is displeased and returns to the BI developer for report changes, repeating the process.

Early-Stage Lessons from Implementing Hadoop

1. Business leaders don’t care how the data is stored or where it comes from. They just want actionable information. And they don’t want to go far to get it.
2. If business users/analysts are given unfettered access to big data sets, it is only a matter of time (usually very little) before your Hadoop infrastructure crashes.
3. Security architectures for the Hadoop/data-lake infrastructures are still being developed and integrated.   Many of these security models are still too new and evolving to be adopted by ancillary vendors.   
4. Because of the lack of security, Hadoop access tends to be granted to a “privileged” few. Traditionally these people are data scientists or programmers.
5. There is frequently a line of end users waiting to get access to these “privileged” few to get access to the data they need

The learnings above lead Hadoop implementers to attempt to re-introduce the data app on big data. What a perfect way to provide wide user distribution while maintaining security and performance for mission critical functions!

However, the important building blocks for developing data driven applications on big data do not exist. Everything must be built from scratch and this is a monstrous task. A developer has to code data connectivity, performance, security, visuals, extensibility, APIs… and that’s just the beginning. Then the developer must adhere to the software development life cycle process for their organization. By the time the data app is built (many months down the road) the time to analytics have STILL suffered - you might as well have returned to the dreaded cycle of pain.

The final leg of the journey is to build a framework that provides everything that a developer requires to build and deploy data driven apps against the organization’s big data infrastructure.

Final Leg Framework

1. Connects to all the data sources you have now or may have in the future, including ones built specifically for your organization
2. Intelligently queries and caches all of your data sources
3. Quickly allows you to join multiple data sources together as needed in a performant way
4. consistently delivers big data to the UI without crashing the browser
5. Authenticates your users via LDAP, Kerberos, or SAML and then appropriately authorizes users for the correct data
6. Provides a complete SDK for developing a front-end
7. Offers APIs and hooks for every phase of your software development lifecycle.

Of course, building frameworks like the one mentioned above can be an extremely complex endeavor and it’s usually a practice geared towards the larger and very technically sophisticated organizations. For organizations that are able, this is the gold standard. However, if you have not built this framework yet-- take a look at Zoomdata. We specifically built a framework to address all of your enterprise data app needs. Give your developers everything they need to build data apps quickly against your big data ecosystem - the time to analytics will be CRUSHED!