WhoZoo: Graph-powered insights into your organization

Using graph modeling to improve collaboration and generate better outcomes

We are living in a world replete with all kinds of networks. Your personal network consists of new and old relationships with people like your colleagues, managers, customers, and even competitors, or with various projects, products, and software. Data about these relationships has become the lifeblood of an enterprise.

While plenty of data about these advanced networks exists, harnessing it is another issue altogether. Finding the right person in a large, distributed organization is a challenge. The old, hierarchical org charts provide few options and quickly go out-of-date. It’s often easy to see who reports to who – but difficult to determine a colleague’s strengths and responsibilities or to understand how people truly collaborate. To make matters worse, outcomes are diminished when employees can’t find the right person to collaborate with, when they need them. Not to mention the tremendous amount of time wasted searching for others, with exploratory emails and phone calls.

A Self-Learning Map

That’s why CA Technologies is funding WhoZoo.io, a lean startup within the CA Accelerator program. This exciting new product is a dynamic, self-learning map to the human beings in dynamic organizations. Using the latest advances in Graph, Deep Learning and Natural Language Processing (NLP) technology, WhoZoo helps connect human beings inside an organization in a meaningful, unbiased and inclusive way. Let’s take a look at how graph modeling can be applied to deliver insights about organizational networks.

Single Relationship Graph Modeling

Figure 1: Graphic representation of a 3-level management chain

Figure 1 shows a graph representation of an organization’s employees who, in this case, are within three management levels from the CEO. In this graph, the blue nodes represent people, and the red lines represent the management hierarchy (who manages whom). Note that only one node at the center of the figure has a self-directed connection, which represents the CEO.

The graph visualizes a snapshot of management chains, as well as the size of departments and teams within the organization. Putting together a series of such snapshots, you are able to understand how resources change over time, as well as the effect of enterprise-level decisions on the workforce. Thanks to graph modeling, a query to produce such graphs can be completed within a few milliseconds, Conversely, constructing similar query using relational data models can prove to be extremely complex to implement and computationally expensive to execute, especially as the depth of traversal increases.

This example illustrates the power of graph modeling in a hierarchical management relationship. Graph modeling enables us to gain insights on multiple dimensions, looking not only at a single entity or relationship, but the entire network with all relationships linking each other.

Multiple Relationship Graph Modeling

Figure 2: Graph representation of multiple relationships between people, GitHub repositories, and programming languages, visualized by Tom Sawyer Perspectives.

Figure 2 shows a graph representation that includes multiple entities and their relationships for software divisions: developers, managers, GitHub repositories and programming languages. Here the blue lines represent the programming languages in which the GitHub repositories are written; the red lines represent who contributes to which repositories; and the green lines represent who owns which repositories. With integrated multi-dimensional relationships in this graph, we can quickly derive many interesting findings that normally would require huge amounts of data across multiple tables in relational models.

For example, we can understand the most popular programming languages used in the enterprise, which may be helpful when deciding the language for developing new products. We can also see what programming languages reside in which repositories and which GitHub repositories have the most individual contributors. Such analysis not only provides insights about the scale of ongoing projects and the potential impact a change would have on users, but also assists in mapping the distribution of contributors from across the entire enterprise, to different ongoing projects.

Advanced analytics and graph technology are at the forefront of the digitally transformed business, improving the customer experience and informing decision making. Such are essential elements to the goal that WhoZoo is working to achieve – enabling more efficient, more collaborative workplaces.

For more information on analytics at CA visit ca.com/analytics or if you’re interested in how advanced analytics can improve workplace collaboration, visit: WhoZoo.io

Cui Lin, PhD, is a Senior Data Scientist at CA Technologies, working on predictive analytics,…



Insights from the app driven world
Subscribe Now >
The Sociology of Software >How (Not) to Lie with Data Visualization >DevOps and Cloud Computing: Exploiting the Synergy for Business Advantage >