David Gay i Tello

Managing two remote teams (12 direct reports) within the Infrastructure Instrumentation organization in Europe.

The Infrastructure organization develops and maintains instrumentation agents and integrations for a wide range of products, including Kubernetes, Cassandra, etc… All the development is open source, usually in Go, and we also contribute to other open source projects such as OpenTelemetry or some Prometheus Exporters.

What I do:

• I improved retention on teams (no employee churn), focusing on strong human relationships where we care personally about each other while challenging directly when needed.
• I invest a lot of time in developing engineers. I coach engineers and deliver regular reviews. Although, I try to act more as a sponsor, looking for opportunities that might help them achieve their goals.
• I consolidated our Software Developer Life Cycle, to ensure the established processes were assumed by engineers and stakeholders, so engineers’ frustration was reduced and productivity increased.
  • Critical and High priority bugs are now fixed within SLA.
  • Reducing Sprint scope, average resolution time has improved 2x.
  • Using software to remove toil, the percentage of Support tickets that become an escalation has decreased by 70%.
  • External contributions are now triaged within a week.
• I set the proper environment, so new projects are built taking testing and reliability carefully into account.
• I work closely with the product and architecture leadership to drive our teams priorities and roadmap. Some examples of teams’ successful launches:
  • New Relic Kubernetes inregration v3, one of the biggest announcements during the last FutureStack’22.
  • New Relic Prometheus Agent.
  • New Prometheus based integrations like IBM MQ or MongoDB.
• I enforce the team to keep some room for innovation and creativity through Innovation Days whose outcomes are shared and land on our roadmaps when appropriate.

Everything and more.

I managed a high performing team of software engineers that integrated cloud services as part of the Infrastructure monitoring, one of the most critical products of New Relic.

The Cloud Integrations team was based in Barcelona. It was partly remote and, as part of the Infrastructure initiative based in San Francisco, had to develop methodologies and communication skills to ensure coordination and alignment across teams.

I was responsible for:

• Maintaining healthy team dynamics and keeping the team output at high levels of velocity, quality and reliability.
• Supporting the team and helping them with high-level tech decisions, technical architecture, as well as to raise blockers and needs from the rest of the organization.:- Partnering closely with product management, engineering managers within Infrastructure organization, support team, tech writers and the rest of stakeholders to set up team priorities and roadmap in order to facilitate project delivery.
• Hiring and on-boarding top engineering talent. I elevated team performance through mentoring and coaching. Attracted and retained software engineers through career path conversations and assigning projects tailored to skill level and interests.

Finally, I worked with the recruiting team and leadership to create a new team, within the Infrastructure initiative, in charge of creating on-premise integrations. I hired and managed first team members until their new manager got hired and on-boarded.

I helped New Relic presenting data using beautiful dashboards and charts that users would want to build with.

I was mainly focused on:

• Designing, creating and maintaining the REST APIs. I started using Python and Django-Rest-Framework, then moving to Ruby and Grape.
• Integrating with New Relic’s microservices arquitecture: API Gateway, authentication, authorization, Feature Flags….
• The improved dashboarding and charting solutions (ReactJS + D3).

Ducksboard was a real-time dashboard that helped you visualize and monitor your data in one single place. Ducksboard was acquired by New Relic in October 2014.

Ducksboard was able to access data from 65 applications, including Facebook, GitHub, Google Analytics, New Relic, salesforce.com, Twilio, Mailchimp and Zendesk. Ducksboard enabled users to show data from multiple external applications alongside internal metrics, and to create dashboards and widgets for visualization through simple API calls.

Ducksboard was used by over 700 businesses and their customers created over 50,000 dashboards.

There, I did everything and more:

• I was in charge of the REST API (Django).
• I spent a lot of time working in the frontend (Javascript + Backbone + D3).
• I was also in charge of the integration with the different Stripe billing APIs.
• Of course, I created and maintained several SaaS integrations (Python + Twisted).

Flumotion is a company located in Barcelona, and part of the Fluendo group. Flumotion is a UNIX-based streaming server, written in Python, on top of GStreamer and Twisted. The platform is a collection of Linux servers distributed over Spain. It handled several Gbit/s of streaming traffic.

Being part of the R&D Team, I was responsible for developing all Web-related technologies in the company. This included:

• A distributed data consolidation system, aggregating +20M daily sessions logged by the Flumotion Streaming Platform per different parameters to feed the final consumption statistics databases in near real-time (Python + Twisted).
• The customer-facing statistics portal, providing fast and highly detailed metrics (Django).
• The internal customer management back office (Django).
• The corporate website and its demosite.
• Customer-oriented web development (in coordination with Professional Services).
• Database(s) design and administration.
• Flumotion Streaming Platfom web-based tools used by Integration & Support Team.

Responsibilities inside the team include:

• Planning and developing the various web projects.
• Bug fixing of the web projects in production.
• Provide documentation to sales and customers.
• Assist in incident analysis and resolution.

During the last months in Flumotion, I participated in a project called Live Transcoding: a new platform which transcoded live streams in realtime into web and mobile formats on the cloud.

Main tasks done here:

• An asynchronous task queue based on Twisted and RabbitMQ. It was focused on real-time operations which include: executing painful tasks written and called by the web app, listening to the new cloud-based platform waiting for any kind of notifications; saving the platform state on database, caching or directly informing the website through HTML5 websockets.
• Customer management back office, including platform API integration (Javascript).

Gextech was a gaming company that provided interactive gaming content to betting and lottery operators in the world.

FantasticLeague was a product oriented to the broadcast television technology that offered live football games every 5 minutes. All games were broadcast live and in real time, each including all the betting markets you could find in a real game.

Tasks I did:

• Functional and technical analysis in order to go from broadcast to internet version.
• Analysis, design and platform implementation (Java + Hibernate + SQL-Server).
• Synchronization system using non-blocking sockets, audio and video streaming.
• Web sites development including video and audio integration (Javascript + Realplayer).
• The platform monitor web application (PHP + MySQL).
• QA support: test plans design, execution and validation.
• Provided technical documentation, integration and sample manuals.
• Supported the integration team based in the UK.
• Product web site development (PHP + MySQL).
• Prototyped a new Gextech streaming solution (Python + GStreamer).

Digital parks was a data center based in Barcelona where I was able to start real web programming.

Main tasks realized:

• Final user support and different providers’ relationship.
• Analysis, design and implementation of the new Digital Parks intranet and all its functional processes (Java Struts).
• New corporate web site development (PHP).
• ERP e-business Suite from Oracle: financial module, order management module and user support module.
• Marketing on-line formation.

I started working as an intern although I was hired quickly six months later, so I was able to finish my studies while working as a developer on PRISMA: a project designed by the Barcelona School of Informatics Computing Laboratory whose aim was to develop an intranet in charge of the UPC studies administration (with Oracle technology).

Technologies I used those early years: Oracle Forms, Oracle Reports, PL/SQL, Java, HTML, CSS…