Michael Dockter Email & Phone Number

West Palm Beach, Fl, Us

West Palm Beach, Fl, Us

Las Vegas, Nv, Us

Las Vegas, Nv, Us

Nubix (https://www.nubix.io) creates software to support Machine Learning at the edge. As an early member of the startup (#3), I was involved in many aspects of startup life. My largest focus was on working with an early customer in adopting the Nubix software.Challenge:Using Alpha and Beta versions of the newly created Nubix software, create an experience for the customer showing the value of the Nubix platform.Action:Create extensions to the Nubix platform that capture SNMP data from various devices, use Machine Learning to create statistical models of "normal" behavior for the devices, and apply the models to detect abnormal behavior in devices.Results:The working system was adopted by the customer. After seeing the results, the customer asked to expand the use of the system to other devices and additional anomaly detection.Skills:Spark, Python, Go, Lua, Machine Learning, Nubix, Terraform, Packer, CockroachDB, SNMP, etcd, serf, raft, Grafana

Armonk, New York, Ny, Us

Highlights of my IBM career:Intellectual Property☞ 28 U.S. Patents☞ 9 Foreign patentsProducts and projects☞ A new, unannounced service on IBM Bluemix catalog☞ IBM Spark-as-a-Service in IBM Bluemix catalog development☞ IBM Optim Performance Monitor on IBM PureApplication systems development☞ IBM DB2 Admin / Object Compare (AOC) modernization development☞ IBM WebSphere Portal Catalog development☞ IBM WebSphere Portal Portlet development☞ IBM Content Management development☞ IBM KnowledgeX☞ IBM Knowledge Utility development☞ IBM AIX Database project☞ IBM DB2 for MVS developmentOdds and ends☞ Presented "Knowledge Utility" to IBM CEO Lou Gerstner☞ Presented "Knowledge Mining Center" to IBM CEO John Akers☞ Presented technology to IBM Combined Board☞ Best speaker at 3rd Annual International DB2 Users Group☞ Best speaker at 18th Mini GUIDE, Florence☞ IBM Top Contributor award☞ IBM Outstanding Innovation awards☞ IBM Outstanding Technical Achievement awards☞ "Lab Director for a Day" at IBM's Silicon Valley Lab

Armonk, New York, Ny, Us

Understanding IBM's IAM.Challenge:IBM's new IAM identity and access management feature was so new there was very little documentation on how to use it. There was no documentation on how to use it end-to-end within software programs. Yet, IBM Bluemix projects were being asked to incorporate IAM into their offerings. Each project team was spending enormous effort to understand the new Attribute-Based Access Control (ABAC) system.Action:Created a working prototype demonstrating how to integrate IBM's new IAM identity and access management feature into offerings on the IBM Bluemix platform. The prototype was designed to help others understand the flow of HTTP requests needed to define and request authorization in the Attribute-Based Access Control (ABAC) system.Results:The results were shared with my team as well as being used by other teams to speed up comprehension of how to program to the IBM IAM protocols.Skills:Attribute-Based Access Control (ABAC) concepts, OpenID concepts, OpenID tokens, JSON Web Tokens, curl, CloudFoundry

Armonk, New York, Ny, Us

Create a "manage" page for IBM Spark service on Bluemix.Challenge:With the advent of more parameters for the user to configure, a new design is needed to the Bluemix "manage" webpage for the Spark service on Bluemix instances. This requires a new visual layout that can accommodate growth and flexibility.Action:To get stakeholder buy-in, I used Balsamiq mock-ups to rapidly iterate through feature and function design of the web page. Next, an HTML mockup was produced to show what the actual page would look like and how the user interaction "felt". Finally, the micro-service that produced the page was created in Node.js.Result:Although the project was accepted, the need for the "manage" page has been postponed.Skills:Balsamiq, Javascript ES6, Bootstrap, Node, Yarn, Gulp, ESLint, Babel, ReactJS, Redux, Webpack

Armonk, New York, Ny, Us

Additional monitoring of Bluemix service.Challenge:The new IBM Spark service on Bluemix is insufficiently monitored. Additional metrics are needed to determine the health of the overall system.Action:I designed and implemented monitoring software for the IBM Bluemix Spark service. Although past efforts were single-purpose code for monitoring only, I created software for monitoring as well as problem determination. The software could be run manually to check system health and isolate problems. It could also be run periodically to create Collectd metrics to be fed to Graphite and ultimately Graphana for visualization.Result:The project went live in December 2016. DevOps also use the program to inspect the system during problem determination.Skills:Go programming language, Collectd, Graphite, Graphana, command-line interface (CLI) subcommand architecture.

Armonk, New York, Ny, Us

Improve internal technical documentation via a "Knowledgebase" repository.Challenge:Internal technical documentation usually gets no attention, but is a valuable resource for future developers who maintain and extend software components. Our internal documentation suffers from the same neglect.Action:Provided "leadership by example" by writing crisp, accurate documentation and encouraging other team members to do the same. I was the number one contributor to the KnowledgeBase repository.Result:Changed the team culture in two ways: 1) Fostered a culture of writing internal documentation 2) Fostered sharing, rather than hiding, informationReduced education costs when bringing on new members.Skills:GitHub markdown, Git, GitHub, Knowledge of the IBM Spark service on Bluemix, the English language.

Armonk, New York, Ny, Us

File transfer over HTTPS.Challenge:In the multi-tenant Spark on IBM Bluemix environment, the secure copy ("scp") command cannot be used to transfer tenant files used by Spark. A new way of transferring files on a per-tenant basis is required.Action: I designed and implemented a "tenant-aware" micro-service to move files to and from the user and the Apache HDFS file system within Spark on IBM Bluemix. This is akin to Amazon Web Service's service that receives HTTP requests from "aws s3 cp ..." commands.Result:Put into production September 2016 to enable IBM Spark on Bluemix's spark-submit.sh command.Skills:Python, Django, Django REST framework, Nginx, Lua, Docker, Make, Maven, Eclipse, logrotate, curl, Swagger, HTTP protocols

Armonk, New York, Ny, Us

Interactive use of Spark service on IBM BluemixChallenge:Until now, the only method of interacting with Spark service on Bluemix is in "batch mode" via spark-submit.sh, which is IBM's version of Apache's spark-submit. Interactive use of Spark is needed for programmatic access to enable programs such as "Jupyter Notebook" to access the Spark service on IBM Bluemix.Action:I designed and implemented a "Spark Interactive API" into the IBM Bluemix Spark offering. The challenge was to take the single-user "Jupyter Kernel Gateway (JKG)" and create multi-tenancy.This was done by intercepting the HTTP requests and WebSocket traffic in the Nginx proxy and forwarding to the appropriate instance of JKG. The code also managed the lifecycle of the JKG instances to minimize the computer resources needed.Result:The feature was publicly available in September 2016. The feature is described at https://console.ng.bluemix.net/docs/services/AnalyticsforApacheSpark/index-gentopic4.html#reference_hbr_wwj_lw. The OpenAPI document is at https://spark.bluemix.net/api-docs/interactive/swagger.yamlSkills:Nginx, Lua, WebSockets, Swagger, CloudFoundry, IBM Enterprise Grade Orchestrator (EGO),

Armonk, New York, Ny, Us

Architect the Nginx proxy configuration for extension and reuse.Challenge:The Nginx proxy configuration's use of "*_by_lua_file" became unwieldy. Every extension was written separately with no reuse. The cost of maintenance and extension rose.Action:I refactored the Nginx Lua code into proper Software Designed Patterns.Result:The code base was reduced by 45%. The Lua code is now easier to extend and maintain.Skills:Nginx, Lua, Software design patterns

Armonk, New York, Ny, Us

"Swiss Army Knife" for driving multi-tenant Spark functionality development.Challenge:Need an internal development tool that drives development of multi-tenancy for Spark in the IBM Bluemix environment. This fosters Test-Driven Development (TDD) where xUnit-type tools are not sufficient.Action:I architected and built a "utility knife" web application that can launch Spark services in CloudFoundry/Bluemix, exercise all of the interior activities, and display the results for verification and debugging. The "knife" was continually updated to drive new multi-tenant Spark features. Result:The "utility knife" shortens the development cycle by simplifying the setup required to do testing beyond normal xUnit testing.Skills:Python, Django, Django REST framework, Docker, Nginx, uWSGI, Make, Maven, Eclipse, CloudFoundry, Bluemix

Armonk, New York, Ny, Us

Startup the IBM Spark-as-a-Service on Bluemix projectChallenge:Using Apache Spark, create a service in IBM Bluemix that can be "mixed" with other Bluemix services.Action:I became the 3rd member of an IBM-internal startup to create IBM Spark-as-a-Service on Bluemix. As a small startup within IBM, the team needed to create the culture and processes for high productivity. An aspect of the challenge was to determine what to "build" and what to "buy". As a team, we tried to leverage as much as we could and build as little as needed.Result:The team grew to about 20 members while maintaining high productivity. IBM Spark-as-a-Service on Bluemix became available in September 2015.Skills:"Startup mentality", Spark, iPython Notebook, ObjectStorage, Git, GitLab, Kanban, Rational Team Concert, Jenkins, Urban Code Deploy, Maven, SoftLayer, Cloud Foundry, Python, Bash

Armonk, New York, Ny, Us

IBM Optim Performance Manager on IBM PureApplication System and IBM PureData SystemChallenge:DB2 databases deployed into IBM PureApplication and IBM PureData need monitoring and managing. IBM's Optim Performance Manager (OPM) does this. To create a more user-friendly experience, integrate OPM into the IBM Puresystem platforms so that databases are automatically monitored and managed.Action:My team and I created the artifacts needed to install and configure OPM at IBM PureApplication / PureData initialization time. We wrote python scripts that ran during lifecycle events for database create, start, update, stop, and delete. We maintained this code through multiple versions of IBM PureApplication , IBM PureData, Optim Performance Monitor, and DB2.Result:The integration shipped with the first version of IBM PureApplication System and IBM PureData System in 2012 and has continued to be integrated into the IBM Pure systems.Skills:Python, IBM PureScale pattern types, IBM Optim Performance Manager, DB2

Armonk, New York, Ny, Us

Increase adoption of "IBM Optim Performance Manager"Challenge:A new product "IBM Optim Performance Manager" (OPM) was created to administrate "DB2 for Linux, Unix, and Windows". Although the product is selling, it is believed that there are opportunities for wider adoption of OPM. Mission: Find those opportunities.Action:Worked with cross-IBM teams to determine where OPM could be bundled into a sale. Also worked with cross-IBM teams to see where OPM could be integrated into product stacks.Result:Deals were set up to incorporate OPM into IBM PureApplication systems and IBM PureData systems.Skills:Perseverance, Following leads, Technical analysis of product architectures and roadmaps

Armonk, New York, Ny, Us

Modernize "DB2 Admin and Object Compare"Challenge:"DB2 Admin and Object Compare" is written in PL/S. Management was noticing that development velocity was slowing down. All new code was custom code and was not able to leverage libraries developed for modern programming languages like Java.Action:After understanding the fundamentals of the situation, I lead a small team that pursued a Java web application based service that an ISPF client would access. This kept the ISPF user-interface that was familiar to the customer while opening up the backing service to modern software development. Result:Although the beta version worked and was tested at customer sites, the decision of management was to continue with the PL/S version. Skills:Java, WebSphere Application Server, DB2, ISPF, XML, Object Oriented principles

Armonk, New York, Ny, Us

Armonk, New York, Ny, Us

IBM WebSphere Portal Portlet developmentIBM Content Management developmentIBM KnowledgeXIBM Knowledge Utility developmentIBM AIX Database projectIBM DB2 for MVS development

Armonk, New York, Ny, Us

Bachelor Of Arts (Ba), Computer Science And Philosophy

Philosophy