Sri Ramalingam’s Post

I was warned that engineering velocity and speed tend to slow down as a startup goes through multiple years of hypergrowth. The real challenge lies in balancing speed, velocity, scalability, quality, and availability—often the unseen factors that support feature delivery—while managing within the investment constraints and cost optimizations. While each stage of a startup’s growth presents unique challenges for engineering teams, we have disproved the myth that growth inevitably leads to a slowdown in engineering momentum. Congratulations to everyone across Harness R&D for a phenomenal year. With 64k pull requests, we have delivered close to 13000 features and enhancements, launched three new modules, our overall platform availability is at a four nine and all of our developer effectiveness metrics are at or above the scale-up industry benchmarks. Delivering these outcomes needs a strong software delivery platform and use of our own modules has helped us to deliver these outcomes. - We have a single CI/CD pipeline with full governance to deploy all of our services and with a 100% availability through the last year. Rollbacks, when needed, are just one click away. 1.5m internal deployments across our environments. - All of our builds runs on our own hosted CI infrastructure. 2.5m internal builds, zero downtime of our build infrastructure and we don’t have dedicated build engineers. Our incremental builds gets done in 30 seconds and the test intelligence saved us about 100k hours of engineering time. - Managing vulnerabilities and library upgrades is a huge head ache for engineering teams. With our Security Test Orchestration integrated into the build pipelines, all of our artifacts gets scanned and if necessary, gets blocked before they reach production. We went an entire year without a single customer escalation due to vulnerabilities. - We are rolling out AI code assist tools internally and the challenge for engineering leaders is measuring the impact of these tools on productivity. Our SEI module tracks a number of code base signals and we are seeing about 30%+ improvement in the commits, PRs and the overall cycle time with these AI tools integrated into the development cycle. Our SEI module became the single source of truth for us to monitor all the code base signals including the Dora metrics.  - We have completely automated our infrastructure provisioning. We can bring up a new production environment anywhere in less than 3 hours with complete observability stack built in and wired into our incident management process. All automated. These outcomes won’t be possible without the entire team having the owner’s mindset, passion, grit and resilience. We are expanding our R&D presence in Bangalore and if you are ready to put the effort and be part of a successful outcome, reach out to Rahul Ravi or Prashant Verma.