When I joined the data stewardship team at LiveRamp, it was ready for growth. The team was quickly learning and evolving, problem solving with ferocity to enhance efficiencies for our innovative customers. However, like many companies, a pile of tech debt stood in our way. We were battling thousands of lines of code across databases and burning more than a hundred hours at a time to deliver excellent customer experiences. It was time for change.
We did what felt both bold and necessary: we slowed down to speed up. The team committed to rebuilding not just our systems, but our team, culture, and processes. We focused on three driving pillars: people, process, and technical excellence.
1. People first: Hiring, culture, and the Ideal Team Player
We had to face a simple truth: we needed more people. Soon, the data stewardship team doubled its engineers from San Francisco to London to Hyderabad. But more than headcount, we focused on mindset. We adopted the Ideal Team Player Framework – Humble, Hungry, and People-Smart – as our north star for hiring and growing talent. This meant:
- Encouraging a feedback-driven, learning culture
- Creating psychological safety through retrospectives and pairing
- Celebrating team wins and helping each other during hard moments
Engineers owned real work from day one, including OKRs and architecture improvements. Junior teammates contributed to technical debt reduction under supervision, accelerating their career growth and pushing the team forward. To support individual growth, we:
- Regularly paired engineers on complex tickets to foster shared learning
- Set up mentorship paths during onboarding and beyond
- Used one-on-ones for regular feedback and career coaching
- Created space for engineers to lead initiatives aligned with their interests and strengths
- Encouraged internal promotions by mapping growth to LiveRamp’s leveling guide
At the same time, we lived LiveRamp’s core values:
- We empower people: Ownership started early, and trust was the default
- We respect people and respect time: We collaborated across timezones while protecting focus time
- We say what we mean and do what we say: We set clear expectations with consistent follow-through
2. Enhancing processes: Making scrum work for us
Next, we standardized our scrum practices while keeping them manageable and human. For example, we implemented:
- Daily and weekly standups based on timezone
- Consistent dev logs, OC tracking, and sprint retrospectives
- Story mapping and T-shirt sizing for OKR planning
Each sprint, we allocated ~20% of our capacity to reducing tech debt. These weren’t side tasks – they were prioritized tasks with real impact. We slightly overcommitted in sprints to push velocity, and used retros to evaluate how much risk was acceptable. We also better aligned with other teams to reduce delivery friction, proactively flagging dependencies. We adjusted planning and communication to fit a distributed team.
Our approach took inspiration from what works on high-performing teams, such as iterative refactoring during active sprints, assigning onboarding engineers to cleanup tasks, and treating architectural improvements as first-class backlog items. But we didn’t just copy and paste. We adapted those ideas to each tool, timeline, and constraint. We applied these priorities to every technique – from commit-to-complete metrics to sprint retros – so it made sense for our team and the context.
Here’s how we put LiveRamp’s values into practice:
- We get stuff done: We adapted, executed, and iterated
- Above all, we do what’s right: We demonstrated sustainable speed over short-term hacks
- We love our customers: Cleaner systems meant fewer bugs and faster delivery
3. Technical excellence: Stability before scale
With better processes and a growing team, we had the resources needed to tackle architecture improvements. We were able to:
- Reduce the codebase by 10% and continue to reduce more
- Migrate to a new tech stack (SingleStore and Temporal) to reduce manual workflows and improve reliability
- Improve PR review quality and modularity standards
- Create dev and staging environments to reduce deployment risk
We also introduced automation to streamline our workflow:
- We began with a manual monitoring process – on-call engineers tracked OC stats daily and shared updates in Slack. This hands-on approach gave us visibility and helped the whole team understand pain points before moving to automation.
- Next, we built Grafana dashboards to automatically track OC metrics.
- We integrated Cursor for smarter code suggestions.
- We cut Docker build and deploy times by 50%.
Lastly, we introduced coding standards tailored to our system. Instead of enforcing a rigid rulebook, we welcomed input and improvements from everyone on the team. This made adoption of the new standards smoother and fostered a sense of ownership. Strengthening our coding standards, along with regular knowledge sharing, helped us build a more resilient, self-sufficient, and high-performing team. Our most-used workflows were redesigned to reduce complexity and on-call burden. We deprecated fragile cron jobs and reduced downstream traffic by 80% while automating internal notifications.
These wins aligned with the following LiveRamp values:
- We love our customers: Our systems have become safer and more scalable
- We respect time: Automation reduced manual tracking and toil
- We say what we mean: We stayed transparent and measured our progress
What changed? The measurable impact
- OC time: 133 → 40 hours/sprint
- Velocity: 11 → 22 points
- OKR Completion: 100% in FY24 Q1/Q2/Q3, and FY25 Q1
- Team growth: 3 engineers promoted, 100% retention
More importantly, customer onboarding today is smoother, the delivery is faster, and the team is proud of how we work.
Next steps for engineering leaders
If you’re facing a mountain of tech debt or have a nascent team, start making improvements with your people and processes. Don’t wait for permission or perfect conditions (they don’t exist).
Here are a few next steps to take:
- Include technical debt considerations in every sprint
- Let onboarding engineers solve tangible problems for customers
- Use dashboards to track OC, velocity, and regressions
- Build a culture around feedback, clarity, and mutual respect
- Adopt Ideal Team Player traits to unlock trust and impact
- Borrow ideas from industry leaders, but tailor them to your team’s needs
Finally, align with your company’s values. At LiveRamp, our six core values aren’t just words on a wall – they helped guide improvements for our team and processes, one sprint at a time.
Building a culture of curiosity and confidence
One of the biggest shifts we’ve seen is how the team approaches the codebase today. Early on, it felt intimidating – even to senior engineers. But things are different now. Through steady cleanup, improved observability, and shared standards, the team engages the code head-on. Engineers are expected to be curious and to actively look for improvement opportunities. We encourage people to pick up challenging tickets, explore unfamiliar areas, and ask deeper questions. As a result, this curiosity-first mindset shows up in code quality, collaboration, and onboarding. New engineers ramp up faster. Senior engineers spend less time unblocking others. And the entire system feels easier to navigate.
We’re not done improving – not even close. But we’re not scared of the code anymore either, and that confidence makes all the difference.