

Secure analytics, insights and intelligence over encrypted data.
Prisma is a solution for data collection, preparation, and visualization for providing insights, analytics, and intelligence from multiple independent data sources in a unified way. Prisma takes care of several inconsistencies in data via thorough data hygiene, which yields a new data structure that favors analysis and many forms of data derivations. Prisma offers a comprehensive set of data visualization mechanisms that can be extended to address the specific needs of various niche applications.
The Problem
When it comes to data-centric applications, the focus is on data as a construction. Therefore, data takes the singular form. But far beyond grammatical concerns, data is indeed singular in its nature, relevance, and possibilities in today’s world. Data is frequently associated with benefits and problems and without the proper treatment, more data will inevitably turn into more problems. Some of these problems are discussed below.
- APIs: Several consumption methods, Insufficient documentation, missing endpoints and attributes, incompatibility with other data sources, overlapping data delivery, are amongst the issues organizations working with multiple APIs deal with on a daily basis.
- Databases: Multiple engines and connection configurations, non-interoperable queries and stored procedures create a state of rigid dependance on each database solution.
- Social media feeds: Datasets obtained from social feeds are highly sensitive to each of their platforms’ choices for attributes and formats.
- Websites: Data is often obtained from techniques such as web crawling and web scraping which can be subject to request limitation, website update that breaks how crawling and scraping works, with no guarantee that the information will be always available.
- Manual and ad-hoc data input: With no systems in place for obtained data, many organizations must deal with manual inputs using spreadsheets, text documents, emails, slide decks, and other everyday tools.
- Lack of hygiene: Dirty data annihilates the utility of valuable data. Data that is duplicated, incomplete, missing, inaccurate, outdated, misplaced, are all examples of lack of data hygiene. Dirty data compromises the efficacy of operations and increases the overall time for executing tasks that depend on that kind of data.
- Lack of Normalization: Different data sources many times contain the same information in several different formats and representations. Some attributes can be also scattered, as if they were independent from each other, when they should be grouped together for context. Unnormalized data can easily lead to duplication and weakening of data analysis via inconsistent information.
- Lack of Standardization: How should financial data be structured? What about inventory, access control logs, web search and navigation history, sales and healthcare records? With every data source owner thinking and acting by themselves, every new integration might require changes in the entire flow of data usage due to how the same type of data is organized and delivered for consumption
- Small subset: When the data available does not tell the whole story, that is, the data available corresponds to a fraction of the information required for some given purpose.
- Out-of-date: It could be hours, days, weeks, months, or even years: if data is not available at the required frequency for any given application, the data served no longer represents the current state of practice.
- Invalid: Policies, standards, laws, specifications, best practices, and other references are all subject to change. That which reflects or is impacted by any of those mechanisms might be deemed invalid if it is not updated when the mechanisms are changed.
Decisions made based on insights obtained from stale data will most likely lead to incorrect results.
- Got data. Now what? Being in possession of valuable data is not enough. An organization might also identify available benefits from proper exploration of the data at hand. Unexplored data has no ultimate practical value even if the value is fundamentally there.
- Missing expertise: Sometimes organizations know what they want from the data they have without actually knowing how to achieve it. In such cases, once again intrinsically valuable data remains useless.
With no proper strategy in place for handling data collection, preparation, processing, and visualization, many companies are flooded with tools with overlapping functionalities while some critical needs remain unattended.
The Solution: Prisma by Algemetric
Prisma is a solution for data collection, preparation, and visualization for providing insights, analytics, and intelligence from multiple independent data sources in a unified way. Prisma takes care of several inconsistencies in data via thorough data hygiene, which yields a new data structure that favors analysis and many forms of data derivations. Prisma provides the flexibility to handle arbitrary formats and complex patterns, ensuring that these datasets are informative. Prisma offers a comprehensive set of data visualization mechanisms that can be extended to address the specific needs of various niche applications.
How Prisma Works
Before Prisma is correctly instantiated, Algemetric conducts an assessment to determine the data-driven needs of an organization. Upon inspection of all the functional requirements and all other specifications from legal and corporate nature, the applicable algorithms are defined on Prisma API and Prisma Dashboard.
Prisma then uses its connector to obtain data from the data sources an organization uses in its operations. Data sources can be from external services and/or from manual input. Once data is properly obtained, it is subject to a comprehensive process of data preparation. The output of the preparation process is sent to Prisma Storage, a unified data structure ready to be used. From that point on, the application is ready to operate via Prisma Dashboard where users will request reports, visualizations, and integrations with external tools when needed.
- Prisma Dashboard: The control room data management. Prisma Dashboard is composed of three main components:
- Data connector: The mechanism responsible for connecting to arbitrarily many data sources and executing a series of pre-defined procedures for data preparation. The Data connector is also able to receive new instructions and expand its data preparation capabilities.
- Data visualization: Every main algorithm specified in Prisma has a visualization mechanism defined for the associated result data set. The default visualization objects are graphs from a wide variety of graph types. However, custom visualization options can easily be added to Prisma via customization.
- Recommendations: From the data it collects and processes, Prisma provides recommendations whenever applicable. Data that is used for recommendations are previously tagged for that purpose so Prisma can use it as input for machine learning algorithms that will provide recommendations according to predefined rules within the scope of the recommendation system in place. Recommendations are typically associated with configurations for digital campaigns, finance management, operations management, supply chain, among others.
- Prisma Storage: The unified data structure organized by Prisma. A clean, ready, fresh, and consistent database for strategic data processing.
- Prisma API: The brain of the entire Prisma data processing operation. Prisma API is responsible for managing algorithms and their output and providing Prisma Dashboard with the information required for its operation.
- Automation: Well-defined procedures must be automated for the sake of time savings and elimination of human error which helps increasing efficiency.
- Organization: When it comes to the data life cycle within a data-centric application, there must be a clear separation between different components, their functions, and the data associated with them. They all must work well together while having distinct responsibilities and scope.
- Complexity reduction: Large data sets from many different data sources working together as input to a wide variety of data processing techniques for the sake of insight generation and supporting strategic decisions handled by a unified and powerful platform. A single solution that eliminates the need for a large number of tools offering partial and overlapping solutions.
- User-friendly: The user experience of data-centric applications must be independent of the engineering complexity associated with them. Such applications must be easy to use, engaging, and rewarding.
- Versatility: The fundamental features of data-centric applications must work as a framework for enabling customized added capabilities to address the particular needs of each customer.
- Augment customer strategic information: No matter how valuable the customer’s data is, data-centric applications must aim to add strategic value to it.
- Data source management: Add, configure, modify, and remove data sources even if no API is available. Prisma also allows for manual input via the Prisma Dashboard. Manual inputs enabled by forms that can be defined and managed also on Prisma Dashboard.
- Report management: An expandable collection of customizable reports.
- Visualization management: Define how data must be displayed by choosing from a collection of predefined visualization objects or creating a new one.
- Integration management: Create and configure an exporting mechanism that transmits results directly to other tools, when desired.
- Recommendation engine: Define the data points that must be considered for data classification and training according to selected decision targets in order to establish a recommendation system.
- Spectra: When working with plain data, Spectra protects data at rest and in transit.
- Obscura: When working with encrypted data, Obscura protects data in use.
- Data collection: Focused on obtaining data inputs, either from external services or from manual input within the application.
- Data preparation: Encompasses all routines associated with data hygiene, unification, and readiness.
- Data processing: Includes all reports and learning on data.
- Data visualization: Involves the rendering of data results with lists, files, graphs, and other visualization objects.
- Built by data enthusiasts: Algemetric recognizes the value of data. We build data-centric technologies because we believe that data is much more than just an input to processes, instead, data has the potential to enable technologies altogether. With this in mind, Prisma is built for taking the most value from data in each stage of the solution.
- Crafted from customers’ insights: We validate the base features of Prisma with our customers and we encourage them to join us in developing, testing, and validating new features in a continuous process of discovery, improvement, and growth, which makes Prisma highly customizable.
- Always protected by Spectra: Spectra is activated on Prisma by default therefore data only moves around encrypted after proper authentication, authorization, and auditing, from and to any component of the application, at all times.
- Available with Obscura: Obscura can be activated on Prisma and on top of the security provided by Spectra, all computations associated with report generation will be executed on encrypted data. With this module activated, Algemetric does not see the data that is being processed, an ideal configuration for higher security and privacy requirements.
- Comprehensive data-centric platform: Turn key Solution that is highly tailored to the needs of the client and encompasses the whole spectrum of the data journey, from Storage to Visualization to Computation.
- Uniqueness: Modular design with a rich set of default features and that can easily be expanded to accommodate additional functionalities and better address particular needs of customers can render a unique instance of Prisma per customer.
- Obscura: When Obscura is activated on Prisma, data is exclusively processed while encrypted. Algemetric never sees your data and therefore data cannot be used for other purposes.
- Spectra: Spectra is activated on Prisma by default for protecting data in transit and at rest. Since Prisma is composed of many different components, Spectra is responsible for ensuring security against the outside world every time data is transmitted from one point to another.
- Sandbox: Every single feature available on Prisma was first part of Prisma Sandbox, a proof of concept mechanism in partnership with our customers for discovering the best possible solution to each new problem they face. Once a candidate solution is identified, tested, and validated by the customer, we deploy it on Prisma.