Robert Berry
OVERVIEW
● IT Manager for the Routematch application and the RMIT team since March of 2020
○ Responsible for the overall security of the Routematch infrastructure to include:
■ The Routematch Application
■ Customer web portals
■ Physical and cloud based infrastructure
■ Internal and edge networking infrastructure
○ Responsible for data center operations and COLO facilities. Working to shut
down the physical infrastructure presence
○ Responsible for IT vendor management. Negotiating pricing and managing
purchase orders as well as reduce spend on Cloud costs
○ Responsible for the management of a cross functional team of System
Engineers, Network Engineers, and Application Engineers
EXPERIENCE AND ACHIEVEMENTS
IT Manager, Software Development (10/2020 – present)
● Security and Technical Debt Context:
○ Security evaluations and security incidents revealed significant security vulnerabilities in
the QTS data center ,as well as, in the Routematch AWS environments. Years of
technical debt and aging systems had led to exploitable security vulnerabilities and
potential outages to customers.
● Contributions:
○ Shored up security in the environments for the interim, and designed a long term solution
to replace the current infrastructure at the time.
■ Hired and managed a strong team that had the skills to mitigate and remediate
the struggles that plagued the Routematch environment.
● Hired a DBA that brought stability to the Routematch database
environment
● Hired a Network Security Engineer that secure the edges of the
infrastructure
● Hired a Network Engineer that brought stability to the network
infrastructure
● Hired a Senior Systems Engineer that allowed us to utilize cloud native
services where possible to reduce technical debt
■ Cleaned up Windows Active Directory environments by conducting a full audit of
the environment and eliminating unused accounts
■ Increased the password complexity for customers and shortened the time for
password changes from never in some cases to every 90 days
■ Identified systems that could be patched safely and brought them up to date on
patches
■ Worked nonstop to reduce the number of CMON tickets that flagged potential
vulnerabilities in the Routematch AWS environment. This number went from over
2000+ to zero over the past year.
■ Directed the strengthening of the edge networking in both the physical and cloud
based infrastructure with the strengthen the “castle wall” proposal, patching and
upgrading network edge devices
■ Proposed and championed the Citrix Cloud solution to replace the on prem Citrix
solutions in both the physical and cloud environments
■ Engaged with AWS engineers discussing best practices that helped to write the
ERD that became the design for the Uber Routematch cloud infrastructure
● Impact:
○ Significant reduction in the number of security incidents, which provided a safer and more
secure environment for customers and their data
○ Provided a path forward to reduce the amount of technical debt in the environment and
secure the customer environments and against potential external attacks and outages
○ Provided a path for the migration of customers from legacy infrastructure to the
significantly more secure Uber AWS environment
● Legacy Migrations Context:
○ After designing and building a more secure AWS cloud based environment to replace the
legacy physical data center and old ATL office infrastructure, we began migrating 190+
customers out of the legacy environment and into the new cloud based infrastructure.
This required an enormous amount of effort and coordination between multiple teams.
After a slow start, we managed to meet our project objectives and close the legacy data
center ahead of schedule.
● Contributions:
○ Identified many complex environments that needed to be migrated from the legacy data
center and office to AWS:
■ The Verizon network used by customer tablets and RMMobile software
■ SureMDM software that is used to manage customer tablets
■ SFTP servers and customer code that was used to manage VLU’s (vehicle
logical units)
■ The “WhereIsMyBus” SMS service that allows riders to get bus locations for fixed
route agencies
■ Various one off services that resided in the old ATL office that had to either be
replaced or eliminated
○ Led the efforts to reverse engineer the legacy “VZW” network in the data center. This was
an isolated network that the customer tablets used to access RMGate proxy and real
servers. It was very complex and took significant time to understand
○ The SureMDM was an old environment that had multiple known security vulnerabilities
and had to be rebuilt from scratch to both remove the list of vulnerabilities and to allow
the hardware team to manage the customer hardware
○ The was legacy Routematch python code that needed to communicate to a legacy SFTP
server in the data center. Had to reverse engineer the SFTP file structure and then had to
build a VLU immulatore to test and ensure the customer VLUs would work with the new
SFTP infrastructure
○ The “WhereIsMyBus” service was driven by a third party application that no longer
existed. Had to recreate the entire by writing a Golang application to replace the legacy
service and have it interact with legacy Routematch services
○ Identified multiple services and legacy databases that were operating out of the old ATL
office environment. Had to engage external third party contractors to help the team
upgrade or eliminate the services so we could shut down the legacy office infrastructure
○ After all customers were migrated out of the physical data center and the office, we
physically dismantled infrastructure and properly disposed of the equipment and released
the data center back to QTS two months ahead of schedule
○ Closed down the InMotion colocation account used by TransitMiner
● Impact:
○ Significantly improved the security posture of the Routematch environment by eliminating
all legacy environments
○ Significantly reduced the number of outages to customers and future outages
○ Reduce to risk of failures due to hardware
○ Eliminated the cost of maintaining a physical data center and the cost of expertise that is
required to maintain one
○ Consolidated all Routematch under a single infrastructure and single infrastructure
vendor
○ By reducing the technical debt, simplifying the environment and improving the security
posture, it better positioned the Routematch application to be purchased by another
company
○ Eliminated the Microsoft licensing that was require for physical environments
● Spend and Vendor Context:
○ Vendor management and vendor spend was out of control in the environment. There
were vendors that no one knew why we used them or why we paid them. Many vendors
could only be paid via credit cards and this was breaking Uber credit card policies and
leading to high credit card expense reports.
● Contributions:
○ Worked with various teams tracking down the reasoning behind particular vendors
○ Once the reason was determined, search for cheaper solutions or alternatives that better
fit the current needs and policies
○ Identified many cloud based services and instances that could be reduced or eliminated
from the Routematch AWS accounts
○ Liaised with the Uber FinOps and AWS Finance teams to clean up a significant number
of past due AWS invoices and automated the process so they would be paid on time
○ Engaged with the FinOps TPM to migrate Routematch Google GCP accounts from credit
cards to being placed under the Uber GCP account
● Impact:
○ Significant reduction in vendor spend
○ Significant reduction to the complexity of vendor management
○ 40+% reduction of physical and cloud infrastructure spend over the past year
UBER EXTRACURRICULAR
Joined Uber Sage ERG
Joined Uber Veterans ERG
