We build Reactive Data Networks (RDN) that synthesize and deliver the latest data to where it is needed most.
Can your web and mobile apps react
to changing data in real-time?
We build Reactive Data Networks (RDN) that synthesize and deliver the latest data to where it is needed most.
Can your web and mobile apps react
to changing data in real-time?
We build Reactive Data Networks (RDN) that synthesize and deliver the latest data to where it is needed most.
Can your web and mobile apps react
to changing data in real-time?
Ziniki Network provides expertise in bridging the gap between the science of composing data and the art of engaging users. We help companies leverage their existing investments in Big Data, and turn these insights into engaging customer-facing experiences with Reactive Data. 

What is Reactive Data?

For a data system to be considered reactive, it must pass these 3 tests:
5339fce0603310f4550000db_Truth%402x.png

Always Reflect the
Truth

If you need to refresh the page on your app or reload the data in your database, then you are not looking at the truth. As the world relies on data to make decisions, the compounding effect of these layers of out-of-date or inconsistent data can cause harm. Reactive Data actively keeps all copies of the data up-to-date in near real-time.

5339fd3faf08d87d2d0000bf_Changes%402x.png

Able to Accept
Changes

The world of Big Data tends to admire beautiful data analysis and visualizations as if they were museum pieces. Most data dashboards take a “look, but don’t touch” approach. Reactive Data must always support read AND write, so the user or the system can affect changes in the moment, triggering updates up and down the chain.

5339fd58af08d87d2d0000c3_Action%402x.png

Can Take
Actions

Traditional data analysis systems can help you make better decisions, but are not usually wired to take actions on your behalf. Similar to how a smart thermostat uses sensor data to keep you comfortable all by itself, Reactive Data must incorporate transactional systems, so that data-driven insights can drive actions and behaviors.

Federating Reactive Data

Reactive Data can come from many sources. By networking any data signals, layering on smart analytics, you can synthesize data into composites to improve existing applications.
5345899cab99f04d280000e2_Four%20Quadrant%20Diagram.svg

Why Build a Reactive Data Network (RDN)?

With an RDN, you can bring together application data from web and mobile endpoints, interaction data from users and things, streaming data from partners, enterprise data from legacy and production systems, and keep everything in sync and reflecting the “truth”.
534449678039b78d17000c29_Enterprise-Large%402x.png

Enterprise Data

Monitor changes in NoSQL, Relational, and Big Data sources and propagate them bi-directionally.

5344497d8039b78d17000c2d_Interaction-Large%402x.png

Interaction Data

Capture user actions and device activities as it happens via Webhooks and JavaScript beacons.

534449878039b78d17000c2f_Application-Large%402x.png

Application Data

Intercept application data access and persistence calls through a REST API layer to capture the latest changes.

5344498e8039b78d17000c31_Partner-Large%402x.png

Partner Data

Consume streaming data from external sources and turn them into easy-to-use and enriched composites.

The Power of Reactive

Reacting to live data to elicit actions in the moment can lead to new possibilities:

Go Beyond Batch Analysis

Current Big Data analytics solutions can efficiently process a large volume of data in parallel. While this is good for forming macro-scale insights, those insights are not immediately useful until the derived data is migrated to user-facing applications. 

With Reactive Data, any insights gained can be immediately pushed to the apps and the connected users.

Break the App Silos

While “there is an app” for almost everything, most of them only do a few things well. To accomplish most real world actions, it is up to end users to act as the go-between these app silos and bring it all together.

Reactive Data can orchestrate the flow of information behind multiple apps that correlates to the actions that end users would have performed.

No More Generic Alerts

Nowadays, users are inundated with alerts, notifications, and various requests. Many of these alerts have little or no informational value and do not facilitate the activities users need to perform.

Reactive Data can identify opportunities to trigger personalized messages that offer targeted ways for users to engage, participate, or transact.

Ziniki Network knows that Reactive Data necessitates a different approach to data integration, beyond what is offered by traditional ETL (Extract, Transform, Load) and EAI (Enterprise Application Integration) tools. By taking an holistic approach to networking data across application boundaries, Ziniki Network delivers on the promise of Reactive Data.
533b210f1d4ceb236300010b_Ziggrid%20Logo%402x.png

Ziggrid: The Reactive Data Engine

Ziggrid is not a database. It is a highly-scalable data processing engine that updates higher-level aggregates in existing databases as lower level signals are added or changed—in real time.
533b248d653a28e83900012d_Polyglot%402x.png

Has Polyglot Support

Can read and write to any database (NoSQL or RDBMS) using existing drivers. Connects to external data sources using REST APIs or WebSockets.

533b24bd653a28e839000138_Functional%402x.png

Is Functional Reactive

Implements a Functional Reactive Programming (FRP) language to decompose complex processing models into a series of simpler steps.

533b24c3653a28e83900013b_Dependency%402x.png

Uses Dependency Graphs

Breaks the silos of stream processing, cube aggregation, batch analysis by chaining calculations or algorithms under an FRP-based dependency graph.

Example: Ziggrid for Baseball Stats Analysis

This demo shows how six years worth of Major League Baseball at bat appearances and win/loss records are streamed into Ziggrid, where a series of increasingly complex analysis is chained together to form a Sabermetrics model. We have slowed down the processing to show how the data propagates through the graph.

In the screencast below, Dr. Gareth Powell, Chief Scientist of Ziniki Network, will step through the various analysis performed and show how the results are continuously pushed to the HTML5-based data visualization dashboard, built using Ember.js, D3.js, and WebSockets.

Watch the full presentation (33 mins)

“NoETL: Processing Data in Real-Time Using Incremental MapReduce”

Presented at Couchbase [SF] 2013 in San Francisco, CA.

Multiple Ziggrid instances can be networked across data centers and organization boundaries.

Ziggrid instances can join a Reactive Data Network (RDN), where changes in one Ziggrid node will automatically propagate to another node that subscribes to that change feed.

This communication is transmitted using our ZINC Protocol (Ziniki Inter-Node Communications Protocol), which uses WebSockets with JSON encodings to inform subscribing Ziggrid nodes, as well as web or mobile applications. ZINC is a bi-directional protocol that can work alongside existing REST-based APIs.

We are working on a simple messaging system that gives human users read/write access to data flowing through the Reactive Data Network.

Ziggram facilitates Reactive Workflows that can span multiple disparate applications and provides an overlay that enables users to share data, edit data, and act on data. This overlay enforces granular security when data is shared between users and beyond organization and domain boundaries. Ziggram messages are traceable, auditable, and preserves lineage.

With Ziggram, different slices of data and application functionalities are brought together in message threads that represents mashups of data, content, and social interactions, elevating the quality of work performed and decisions made.

We Love Open Source

The people behind Ziniki Network have helped develop several open source libraries:
533cde482d82e8116c000814_Ziggrid-Logo-on-Light-Background.png

Ziggrid Language Runtime

We have open sourced the core language specifications of Ziggrid, so that we can work with a broad spectrum of data scientists to define the language features and standard functions that can be used to describe data relationships and calculation in a functional reactive way. We build the first version Ziggrid on top of an open source NoSQL data store called Couchbase, and are in the process of porting Ziggrid engine to run on top of the in-memory processing capabilities of Apache Spark.

533ce08d2d82e8116c00081f_CardStack-Official-Logo.png

CardStack

We are working on a new way to build reusable Reactive User Experiences that work across a variety of devices and operating systems by adopting the “card” metaphor and supporting it with a few key runtime capabilities via JavaScript. We plan to consolidate a few existing open source projects, including Oasis.js for card sandboxing, Conductor.js for card contracts, and Glazier for card assembly and user security, to create an easy-to-use developer experience for “card” development.

We are active users of the following open source projects and are committed to contributing back to the community:
533ce261a370709d400008e6_emberjs-tomster.png

Ember.js

We love the Ember.js community and believe that Ember will become the “go-to” front-end JavaScript framework for ambitious web applications when the dust settles. Ziniki has contributed to the architectural design of the Ember Data library to give it the right semantics to support streaming change sets from the cloud. 

533ce472a370709d4000095a_Couchbase%402x.png

Couchbase Server

We use Couchbase Service as both a transactional (OLTP) and analytics (OLAP) data store. Couchbase is scalable, document-oriented NoSQL engine that provides the foundational persistence and data processing layer for the Ziggrid engine. 

533f39fb5833c2ac49000035_Cloudify%402x.png

Cloudify

We want to allow a Ziggrid cluster to horizontally scale out as data processing volume fluctuates in the Reactive Data Network. Cloudify allows us to develop run-time “recipes” that monitor the usage patterns of the virtual servers on Amazon Web Services or OpenStack environment in real-time, and allow to us to “react” to spikes in data traffic by bringing up (or down) additional nodes for Ziggrid cluster or the underlying data store cluster.

Helping You Get Started with Reactive

We like to work side-by-side with our clients and found the following steps to be the most successful:
PHASE 1

Reactive Strategy

Timeframe:

  • Up to a week

Work Products:

  • Product Concepts
  • Data Scoping
  • Strategic Roadmap
PHASE 2

Reactive Design

Timeframe:

  • 2 to 6 weeks

Work Products:

  • Design Mockups 
  • Technical Architecture
  • Feature Prioritization
PHASE 3

Reactive Development

Timeframe:

  • 4 to 12 weeks

Work Products:

  • HTML5 Interactions 
  • Data Source Integration
  • Algorithms & Metrics

What We’ve Been Up To

Our network of technologists, designers, data scientists, and product strategists have been working with select partners on a wide range of Reactive Data applications. The following examples are a select list of of our current engagements:

Reactive
Learning Environment

Helped a large education publisher to design an open ecosystems-based course building and delivery platform, using data to react to student needs.

533d53b92d82e8116c000ec8_Four-Quadrant-Diagram-Small
  • Online Course App
  •  Standards / Benchmarks
  • Learner Activities
  • Institution SIS / SSO

Watch a presentation on EdSense

Reactive 
Sports Betting

Working with a UK-based gaming firm to envision the future of mobile betting using personalized offers, such as real-time buybacks.

533d53b92d82e8116c000ec8_Four-Quadrant-Diagram-Small
  • Mobile Betting Apps
  • Real-time Game Feeds
  • Previous Betting Habits
  • Operator CRM / Book

See our one-pager on Reactive Betting

Reactive
Customer Self-Service

Strategizing with large call center operator to improve customer self-service by providing live data syncing between agent and customer web apps.

533d53b92d82e8116c000ec8_Four-Quadrant-Diagram-Small
  • Customer Live Form App
  • Vendor Pricing Lookup
  • Prior Customer Actions
  • Siebel CRM for CSOs

Reactive Human Performance Tracking

Strategic discussions with a D.C. based firm and the DoD on a new way to continually assess force readiness and quantify operational performance.

533d53b92d82e8116c000ec8_Four-Quadrant-Diagram-Small
  • Environmental Overlays
  • Sociological Patterns
  • Quantified Device Data
  • Normative Benchmarks

Reactive 
Patient Care

Strategic discussions with a large hospital to push patient data with personal care checklists directly to the providers’ devices throughout the visit.

533d53b92d82e8116c000ec8_Four-Quadrant-Diagram-Small
  • Care Provider Devices
  • Health KPM / Thresholds
  • Medical Device Data & Tests
  • Patient Care History

Reactive
Analyst Workflow

Working with a D.C.-based systems integrator to optimize an agency’s upstream data collection, so that the data can be acted on, edited and traced.

533d53b92d82e8116c000ec8_Four-Quadrant-Diagram-Small
  • Analyst Workspace
  • Data Mining Results
  • Workflow Interactions
  • Exisiting Knowledgebase

Let’s talk more. Call us at (971) 71-REACT.

We would love to discuss how Reactive Data can transform your business. Give us a call at +1 (971) 717-3228, send us an e-mail at info@zinikinetwork.com, or fill out the contact form below:

Thank you! Your submission has been received!

Oops! Something went wrong while submitting the form :(

The Ziniki Network team is conveniently located in the following metro areas:

Washington, D.C.
533db039c51278ef42000111_Location-DC.jpg
New York City
533dadd44b5c62f0420000cc_Location-NYC.jpg
Philadelphia
533db041c51278ef42000113_Location-Philly.jpg
San Francisco
533db075c51278ef4200011c_Location-SF.jpg