Numerify: A System of Intelligence to Support Data-Driven Strategies to Optimize IT Performance
Product Mission
Numerify has broken new ground by creating a product that implements a “System of Intelligence” that delivers AI-powered analytics for IT performance management. In essence, Numerify has productized a sophisticated form of business intelligence for IT.
While the mission of IT has been to promote the use of technology and data through applications and analytics to improve business results, the practice of IT has often not been as data-driven as the rest of the business. Given how many systems are now in use to manage the various processes in IT, the time is ripe for the application of metrics and expanded use of data to track things like:
- Operational processes used to run IT
- Timeliness and effectiveness of projects delivered
- The increasing number of partner and outsourcing relationships
- Quality and business value of service delivery to users
The mission of Numerify is to allow an IT department to:
- Connect to all the data tracking IT-related activities across Plan, Dev, and Ops
- Configure a set of standard analytic applications to create the KPIs and analytics dashboards needed to run IT in a data-driven manner
- Demonstrate through business metrics the value IT is delivering to its business stakeholders
- Gradually expand the scope of activities in IT being run and optimized using data-driven methods, and, in the process, create a dynamic and agile culture that uses metrics to drive continuous improvement.
Origin Story
Numerify was founded by three ex-Oracle and one ex-IBM employee who realized that the migration to the cloud (The Great Migration) and the fragmentation of systems caused by the rise of SaaS software (The Great Fragmentation) were going to lead to a crisis in analytics and BI. The rising number of applications and data sources would be difficult to connect to and aggregate into a coherent view of business activity. Numerify was founded to attack this problem by creating a System of Intelligence with IT being the first area of focus.
Numerify was co-founded in 2012 by Gaurav Rewari, Srikant Gokulnatha, Sadanand Sahasrabudhe and Pawan Rewari, who drew on their experience at Oracle and IBM to recognize the need for productized analytics in the cloud. Based on their prior experience, Numerify’s co-founders recognized two major trends occurring at the time: 1) a great migration to the cloud and 2) a great fragmentation in terms of products and vendors due to that migration. Companies of all sizes were migrating to the cloud and that trend was sure to continue. But as it did, the number of apps and products companies were using was greatly increasing. The Numerify team recognized that once this occurred, and companies automated processes in the cloud and implemented new apps, they would want readily available insights from their data. But because of the fragmentation, data was captured across many apps and platforms, both in the cloud and on-premises. The Numerify team thus recognized that there was going to be a resurgent need for unified insight across this newly fragmented landscape of cloud and on-premises apps: their premise was that companies migrating to the cloud would rather buy packaged analytics than build custom analytics. The company’s mission became building productized analytics that addressed this fragmentation.
Based on what they saw at Oracle and IBM, Numerify’s co-founders saw there was definitely an appetite from customers for best practice analytics delivered as a packaged offering, rather than reinventing the wheel with custom models. But they also saw the underlying systems of record were moving to the cloud and getting more diverse, and so offering similar solutions in the cloud, as a cloud-based offering, was an opportunity.
Once they started to build a platform for productized analytics in the cloud, they realized that IT was underserved when it came to analytics and decided to make this the target domain for Numerify. They quickly realized that in many senses IT is the cobbler’s children when it comes to BI and analytics technology. For decades, although IT had spent huge amounts of time implementing systems that instrumented the business in great detail and provided key metrics that guided management and various other types of operational decision making. But for IT itself, the use of BI was extremely limited.
This was doubly ironic, not only because IT was not eating its own dog food, but because IT is one of the most information-rich areas of most companies. The amount of data that is generated from even the most simple IT system is immense. This data can be used for diagnostic purposes, to create predictive models, to capture and provide alerts about important events, as well as to perform root-cause analysis of problems. Additionally, because IT ran nearly the same in all verticals and industries, productized analytics for IT could be universally applied.
In addition, there is a broad commonality of vendors across the IT landscape. IT management practices, while not standardized, have been converging in ways that have been influenced by broad design and management trends (such as Agile, DevOps and its children such as continuous integration, continuous deployment, and new variants such as AIOps). IT management frameworks (such as ITIL and COBIT) have also influenced the way many companies are run and caused practices such as change control boards to be widely used. The data landscape was disparate enough that visibility was a challenge, but the commonality of standards and approaches was high enough that the problem was solvable and broadly applicable.
The focus of the vendors on specific niches also provided an opportunity. The in-app reporting of cloud-based systems of record was rudimentary, typically limited to operational, here-and-now reporting, or simple KPI trending, similar to their on-premise predecessors. This made sense because their technology stack was geared towards supporting multi-tenant or multi-instance operational workloads rather than analytical ones.
Gaurav Rewari saw that this vast landscape of data pouring from IT systems, and the new understanding of how to manage and run IT that could be gained from it, offered an arena ripe for productization. By creating semantically aware connectors to all of the sources of information, implementing an extensible data warehouse to contain that information, and then creating a common set of metrics, data visualizations, predictive AI/ML models, and dashboards that were used to put in place common management and analysis practices, an advanced product for IT Business Analytics could be offered out of the box.
That’s exactly what Numerify has done with the core functions for IT management, which are being steadily expanded to include more domains as they are defined and understood. As Numerify has grown, its approach has been to create solutions that are less costly to implement than custom builds while also the speeding time to value companies experience. They follow the 80/20 rule, in which they try to provide 80% of what a company needs as fast as possible to provide faster time to value and give them the ability to fill in the remainder over time.
Pain Points and Intended Users
Numerify helps to address the pain points of people throughout an organization, from the C-suite on down. Here is a summary of the problems Numerify helps address for each role.
Numerify Roles and Pain Points
Role | Pain Point |
---|---|
CIOs | * Cannot understand the quality of IT operations and service delivery, both at the level of each system and for the organization as a whole. * Cannot understand and track progress toward improving service delivery and reducing systemic problems. * Lacks data to make key decisions about simplifying and pruning the IT portfolio. * Finds it difficult to justify large investments that would improve the systemic, long-term performance of IT. * Needs to move money and resources from keeping the lights on to innovation * Needs to drive app rationalization and cloud migration initiatives, determining which apps and workloads to prune or move to cloud first, and why * Must minimize risk of production failures * Needs to measure progress of transformation initiatives, such as charting and tracking the company's journey from waterfall to agile/DevOps |
IT VPs or Department Heads | * Lack of visibility into current state; lacks an end-to-end view * Overwhelmed by the huge number of urgent problems facing their staff * Lack of information to prioritize projects * Finds it difficult to identify root causes of problems * Unable to provide models that show the business value of IT investments that they operate |
IT Staff | * Lacks detailed information about the operation of each system, presented in a coherent summary * Has difficulty managing partner relationships to ensure proper service is delivered and that gaming of the system to mask problems is avoided * Cannot quickly identify specific problems across all channels that are affecting key business metrics * Lacks ability to diagnose and drill down to get more detail |
Line of Business Managers/ Senior Executives | * Lacks detailed information about what IT systems are doing for the business and how business metrics are affected by IT quality * Lacks an understanding of the business value IT is providing * Has little visibility into improvements in IT service delivery and quality over time |
Analysts/Data Scientists | * Has not understood the context for applying business analytics to IT because there were a dearth of metrics. * No access to organized and integrated data that forms the basis of applying analytics and creating predictive models * Wants to ask the next set of questions as they arise |
Partners/Outsourcing/ Service Providers | * Has found it difficult to provide comprehensive information so that the quality and productivity of their services could be improved over time * Has tended to manage focusing on terms of the SLA instead of providing greater business value to IT |
Elevator Pitch
Numerify’s System of Intelligence for IT integrates all of the data needed to plan, build, and run IT using a common data model with productized data integrations and analytics that provide key solutions for the top areas that are crucial to IT management.
Numerify’s KPIs, AI-powered analytics, and dashboards are focused on five key areas:
- IT Service analytics
- Project and Portfolio analytics
- Software Development analytics
- IT Asset analytics
- IT Contact analytics
The KPIs, analytics, and dashboards in each area represent a perspective on how IT leaders can easily and quickly answer the most pressing questions.
Numerify offers an 80/20 solution in which companies get most of what they need from the platform to ensure greater results and time to value than they would with a custom-built solution. Using their experience with many IT clients, Numerify can help to manage the rollout of the platform, the use of analytics, and the implementation of best practices.
Numerify applications can be purchased a la carte, but are interconnected by design and able to support solutions that have a larger scope than covered by any one application, as explained in the Solutions to IT Challenges section.
Central Dogmas of Numerify
The central dogmas of a product are the fundamental assumptions that guide its creation and help inform business and product strategy, product management, and engineering decisions. Here are the central dogmas at the foundation of Numerify:
- IT departments have a great opportunity to take advantage of BI and advanced analytics to improve their own operations and run themselves like a business.
- Productizing KPIs, analytics, dashboards, and AI through a System of Intelligence for IT is essential for digital transformation of not only the IT function but also the entire business.
- A productized System of Intelligence can accelerate time to value by providing pre-built data integration, visibility, and analytics capabilities that are needed to address the most common IT operational challenges. This is especially needed in the fragmented environment of cloud applications.
- The information provided by a system of intelligence can be used to improve a wide variety of operations by bringing broad awareness of key metrics to “plan, build, run” cycles.
- A System of Intelligence must solve as many common integration and implementation challenges as possible and at the same time provide extensibility, autonomy, and self-service.
- A general purpose System of Intelligence should support a generic process including comprehensive analytic capabilities that ascend in power through the following four stages of maturity: Visibility, Investigation, Correlation, Prediction.
- Cloud-style automation must be brought to bear to make it easier to run a System of Intelligence. The construction, configuration, deployment, and operations of the system is a complex problem made harder because the data sources frequently change. A traditional approach based on addressing this complexity with human services or brittle integration code breaks down in the world of the cloud.
- Business needs and therefore analytical needs change and evolve. An analytical solution needs to be agile and nimble to respond to these changing needs.
- Analytics serve business needs. A technology/feature driven approach to analytics will rarely deliver sustainable business value. On the other hand, a domain-specific analytical solution that is focused on covering a range of questions will be more effective.
- A SaaS solution complemented by managed services is a great fit for IT departments that are hard-pressed for resources, yet need analytics quickly to respond to the needs of digital transformation.
- Delivering a System of Intelligence on a platform allows existing solutions to be upgraded with new functionality while preserving customizations and extensions.
- A platform-based System of Intelligence enables the customer to leverage “common core” best practice content and also customize to meet their specific needs.
- The platform unifies the IT PMO, Development, and Operations teams around shared metrics that are undisputed and measure the most important performance goals of the organization, which helps shape behavior, drive long term decision making and increase alignment.
- A platform approach ensures faster time to value than companies would experience with a custom solution.
- A platform enables companies to experience deep, wide, and rich analytics at a much more cost-effective price point than with a custom solution,
- A platform should work well with existing customer processes, custom applications, and preferred BI tools whether they are on-premises or in the cloud.
Product Capabilities
Numerify’s product comes to life in three ways that build on each other:
- A platform built using a declarative approach to allow the data integration, modeling, and delivery of analytics to be performed in an abstract way, as independently as possible from any specific implementation.
- Applications built on the platform to assemble the KPIs, analytics, AI and ML, and dashboards needed to create systems of intelligence for specific IT use cases.
- Solutions that are tailored from the applications to solve broad, crucial, and enduring problems facing CIOs.
This architecture delivers perhaps the most important goal for Numerify: the ability to deliver an out of the box product that meets 80% of customer needs but also allows customization in a manner that does not get in the way of later upgrades to both the platform and packaged application components.
In this way, Numerify has created a product that adheres to the company’s central dogmas.
The Numerify Platform
The Numerify platform, which provides the foundation on which Systems of Intelligence are created, has three key jobs:
- First, the platform must be a continuous pipeline that collects, integrates, and normalizes data from all relevant IT systems and makes it available for use by analytics and applications.
- Second, the platform must provide the plumbing and tools needed to create analytics and applications out of the integrated data.
- Third, the platform must allow customizations so that both customers and partners can extend existing applications and build new ones.
This diagram shows the high-level structure of all of the moving parts in the Numerify Platform:
Declarative Data Transformation and Integration
Numerify’s approach to the first job, data transformation and integration, represents one of the most powerful features of the platform.
Numerify’s founders realized that to create a future-proof product, to accommodate customization, and to speed development, it was essential to use powerful and elegant computer science methods that involved solving the problem at the highest possible level of abstraction, but in a way that allowed the heavy lifting to be automated.
To accomplish this, Numerify created its own data transformation language using a declarative approach, which separates the specification of the transformations from the implementation. Essentially, Numerify has created a domain-specific language for data transformations of IT-related data.
This declarative language allows Numerify to shape the data in layers to meet the needs the needs of each application. Once the data is in the right format, an analytics layer is used to answer questions and deliver those answers in the consumption model preferred by the end-user.
The declarative approach not only simplifies integration and makes it more maintainable, but has benefits for the entire product lifecycle, making upgrades easier, making automation simpler and more powerful, and allowing technology refresh with minimum impact.
Because the language is declarative, it is separated from implementation, allowing Numerify to delivery a robust semantic integration of data that is independent to a large degree of underlying execution technology. This allows changes in the underlying connectors and data to take place without requiring changes to the layers above it. In addition, Numerify can replace low-level implementation technology as better solutions become available.
For the end users, this means that their analytics are both future proof and durable through major events such as acquisitions and mergers, when new IT management products may arrive. The future-proof quality is delivered because Numerify supports all IT management systems. So if you replace BMC, ServiceNow, or CA with a competitor, you can just start absorbing the data from the new product into Numerify. In a merger or acquisition, it is likely that any new product introduced into the landscape will already be supported.
Analytics and Application Development
The second and third jobs of the platform are to develop analytics on integrated data and to present those analytics to end users in ways that support the goals of each application. This layer of Numerify is less abstract. The platform uses a variety of analytics and application development tools to create applications that can be controlled both through configuration and by writing code.
At this point in the development of Numerify, most app customizations are performed by Numerify professional services. As the platform matures, the development environment will first be made available to partners and eventually to customers.
Hosting and Infrastructure
Numerify was born in the cloud, embodies cloud style-automation and simplification at every level, and runs as a SaaS application. Numerify is certified compliant with common standards such as SOC2.
Here is how each layer of the platform works:
Layer: Data Connectors
Numerify connects to all of the products commonly used to manage IT. Access to data sources is usually gained through an API., The challenge is to bring the data in and assemble it into objects with common semantics. For example, data from any number of IT Service Management systems can all find its way to one type of object such as a problem.
This semantic integration starts at the connector layer, which uses declarative transformations once the data is retrieved, and continues in the data integration layer.
In addition, the connectors play a role in ensuring the operational quality of Numerify. Data sources are profiled and undergo data quality analysis, so if something goes wrong at the source, it is quickly identify before bad data is ingested into Numerify.
Key points:
- Numerify has a wide array of connectors to all commonly used products for the plan, dev, ops cycle including those by CA, Atlassian, AppDynamics, MicroFocus, ServiceNow, and BMC.
- Data from on-premises products and data sources is collected and pushed to the cloud using software called the Numerify Data Collector that runs behind the firewall in the client environment.
- Numerify’s declarative language is used to do data transformation so that changes to the APIs and underlying data model have minimal impact.
- Numerify’s semantic models (more on these below) are data source agnostic, so Numerify can integrate with any system of record used to manage IT, whether it resides on premises or in the cloud, even if it is proprietary and built in-house.
- Numerify can create custom connectors for data from proprietary systems as a services engagement. The product roadmap includes adding a developer toolkit for connectors that can be used by customers and services/technology partners.
- Data profiling and quality control checks are used at the connector layer to detect schema changes and data quality issues.
Layer: Data Integration and Transformation
In addition, the data is put into a common semantic model that is then shaped to the needs of each of the Numerify analytic applications. Numerify has created a common data model that represents the semantics of all of the data used to run IT. The data model can be extended to incorporate proprietary elements of an individual customer environment.
In this way, data that starts out in a distributed, heterogeneous form ends up in a semantically unified form to support the creation of KPIs, analytics, and dashboards and use by AI/ML algorithms.
This transformation language is used at the connector layer but is most widely and deeply employed in the data integration and transformation layer to achieve three types of data integration:
- Level 1: Technical — Gain access to the data whether through a database query, a file, or an API.
- Level 2: Metadata — Understand the semantics of what each table, column, entity, and attribute means as well as inheritance relationships.
- Level 3: Use in Practice — Understand how the data is used in practice and how it has been extended and customized.
Layer: Data Repository
Numerify uses high performance, cloud-based, data lake and columnar data warehouse stores as its central repositories, along with other utility repositories housed in operational SQL databases.
Key points:
- Numerify stores historical data in this repository, so if a solution is switched out, the history is not lost.
- Access to the repository is fully abstracted so it may be updated at a future date to take advantage of the best data warehousing technology that becomes available.
- The columnar structure of the data warehouse greatly speeds queries.
Layer: Analytic Application Framework
The data integration layer provides access to data objects with associated metadata and data lineage that have been crafted to support the KPIs, analytics, dashboards, and AI/ML use cases for each application.
The general structure of each Numerify application is to support the following phases used to resolve problems and optimize operations:
- Visibility
- Investigation
- Correlation
- Prediction
For each of these layers, applications are created using a full-featured BI and analytics development system used as an embedded system.
In addition, the data model is accessible so that any of the common BI tools that customers may already have in place can also be used. This is known as the Bring Your Own BI method (BYOBI).
The key task of the analytics layer is to:
- Provide the capabilities needed to create applications.
- Provide the analytics foundation that supports the UX layer’s ability to allow end users to interact with the application using the consumption model of their choice, including reports, interactive dashboards, their preferred BI and analytics tools, etc.
Layer: UX, customization and extensibility
Numerify’s UX layer is designed to allow end-users to explore the data in the ways that they find most natural and beneficial, including:
- Curated scorecards,
- Interactive analytics,
- Ad-hoc analytics, and
- Advanced analytics (like AI/ML)
The UX is configurable and can be extended through professional services engagement.
Benefits of the platform:
- Because the system of intelligence is based on a platform, the system of intelligence can be configured to meet the needs of a specific company in a way that allows the platform to be upgraded.
- Numerify thus helps companies understand their source data at a deeper level.
- Numerify leverages industry best practices and experience with many customers to inform the design of data models, KPIs, dashboards, and analytics so customers ask and answer right questions.
- The platform that is built to change and support customization: The architecture makes it easy to not only rapidly build new applications, but support functionality upgrades with no customer effort.
- A platform approach offers a standard solution across the entire enterprise. This allows insight whether teams are using Jira, TFS, ServiceNow, or other common IT management tools.
- When a customer implements new software implemented, Numerify can easily integrate with these because all common vendors are supported and the data models are source agnostic.
Applications
Numerify’s platform truly comes to life and creates value when it is used in the context of specific applications and solutions. The final step in productization happens in the application, where all layers are knitted together to serve the needs of an IT professional who is doing a complex task.
Here is a summary of the applications currently in the Numerify portfolio.
Software Development Analytics
Software Development Analytics (SDA) improves an application team’s efficiency and quality across the SDLC process. SDA helps users gain immediate visibility into SDLC processes by leveraging critical data trapped in development and testing tools. Understand how Build and Run processes impact release velocity to make optimal decisions throughout the application life cycle. Gain new insights by blending and modeling data across multiple IT systems of record. Understand the full impact of application quality issues on production and the business by analyzing SDLC and operational data together in one place.
Project and Portfolio Analytics
Project and Portfolio Analytics (PPA) helps IT leaders deliver projects on time and on budget with multiple views of project costs, progress, activities, and resource usage. It is a complete analytical application that provides full visibility into project activities, from the portfolio level down to individual time cards. The application leverages data from typical project management and development systems like CA Clarity, Microsoft TFS, and Atlassian JIRA.
IT Asset Analytics
IT Asset Analytics (ITAA) helps IT leaders clearly understand the state of their assets and solve asset-related challenges. The ITAA application improves IT asset management by uniting asset data with interactive dashboards and Numerify’s analytical domain expertise. Analyze data from the CMDB as well as other systems such as ITSM and financials to improve asset performance and effectiveness.
IT Service Analytics
Numerify IT Service Analytics (ITSA) helps ITSM leaders tackle complex processes and drive continual service improvements. ITSA converts data from service management processes into key insights. Move beyond simple process KPIs to analyze deep audit logs and complex data sets. Extract unstructured and highly granular data from the CMDB, service catalog, knowledge management system, and other relevant systems of record. The solution serves data-driven IT service management and operations professionals alike.
IT Contact Analytics
Contact Analytics (CA) helps contact center teams increase productivity and accelerate operational improvements by identifying drivers of customer contact, agent performance, and business impact. Obtain a complete picture of operations by combining relevant data sources, including pre-built integrations to popular ticket systems and automatic call distributor systems (ACDs). Improve visibility at every level to help drive alignment and enable change across teams for optimal performance.
Change Analytics
Change Analytics is a Machine Learning driven approach to change success. Gain the intelligence needed to reduce expensive outages and increase change velocity. Leverage purpose-built Machine Learning models and analytics capabilities to predict risky changes and identify actions that will reduce change failure. Identify risky changes and proactively take steps to reduce risk or prepare immediate remediation. Investigate and uncover systemic causes of change failure that span people, processes, and technology. Leverage machine learning models to reduce change failure rates and associated incident MTTR
Solutions to IT Challenges
Numerify’s large and diverse client base means that it acts as a clearinghouse for the experience of the IT departments. There are patterns across their customers that map very nicely to various digital transformation initiatives that most businesses are now undertaking.
Numerify puts this knowledge to work in its products, but also in solutions that are tailored applications designed to solve crucial problems that are often idiosyncratic. Unlike Numerify applications, which are precisely designed to meet a specific need, solutions are combined by knitting together multiple Numerify applications and adding additional capabilities.
In general, Numerify solutions increase management visibility and transparency across the plan, build, and run process and enable accelerated diagnosis of problems and a faster cycle of optimization.
Here are the four areas for which Numerify currently offers solutions:
Agile Innovation
The Numerify Agile Innovation solution helps development organizations adopt and continuously improve practices like Agile, CI/CD, and DevOps. Gain a single pane of glass across the entire development lifecycle spanning Plan, Build, Integrate, Deploy, Operate, and Feedback. Make data-driven decisions that enable improved developer productivity and code quality.
- Gain full visibility across the SDLC for each project, showing project momentum, probable outcomes, and optimal times to exert control.
- Make continuous improvements in Agile, CI/CD and DevOps adoption, through a cycle of identification, removal and re-evaluation of key constraints.
- Improve communication within the development team as well as with external stakeholders like business sponsors and IT operations.
Application Health & Modernization
The Numerify Application Health & Modernization solution provides key IT stakeholders with the shared intelligence needed to make effective rationalization, modernization, and operational decisions that continuously deliver the right portfolio of applications aligned with business needs. Gain common unified strategic views across siloed metrics to shape the portfolio with broad visibility across multiple applications. Make data-driven operational decisions based on composite application health visibility and scores across performance and IT service metrics such as incidents, problems, changes, outages, downtime, user and business impact.
- Application Portfolio Optimization: How do I enable stakeholders from IT, business, and the program office to make timely investment, rationalization and strategic decisions?
- Application Health Optimization: How do I prioritize remediation projects that have the greatest customer impact?
Service Excellence
The Numerify Service Excellence solution provides organizations with the intelligence needed to efficiently deliver innovative and responsive IT services. Gain a single pane of glass and break down silos across service delivery teams and processes like Service, Incident, Change, Financial, and Human Capital Management. Make data-driven decisions on how to balance your organization’s most essential resources (employees, investment, and innovation) as follows:
- Service Cost Reduction: How do I maintain or lower costs while keeping the CSAT scores of my customers high?
- Service Experience: How can I deliver services that will keep employees productive, engaged, and innovative?
- Service Provider Performance: How can I leverage external service providers to deliver innovative services while keeping them accountable?
Resource Planning & Allocation
The Numerify Resource Planning & Allocation solution provides the intelligence needed to be an agile IT organization that continuously adapts to corporate objectives. Make data-driven decisions on how to allocate resources across project and service delivery. Accurately forecast resource needs and effectively make the case for additional resources as follows:
- Use Machine Learning model powered analysis to predict resource shortages or overages and plan allocations to meet project and service demand
- Identify opportunities to augment staff with external provider(s) based on historical performance and the expected value of supplied resources
- Continuously measure resource “work time” to pinpoint bottlenecks, increase operational efficiencies and change velocity
- Bridge the gaps between corporate strategy, project portfolio, and resource allocation; redirect resources to higher-value projects and increase execution velocity
- Better utilize resources through up-to-date head counts and work capacity metrics