Project Name
Scaled a Series B Startup From 5 to 50 Engineers Without Losing Deployment Velocity, Restoring 10x/Day Frequency With Backstage & Dagger on OCI
Our client is a Series B technology startup based in North America, building a SaaS platform entirely on Oracle Cloud Infrastructure. Core services run on OCI Kubernetes Engine, with event-driven workloads on OCI Functions and transactional data on Oracle Autonomous Database.
At 5 engineers, the founding team shipped daily – 10-plus deployments per day was the norm. As the company closed its Series B and headcount scaled from 5 to 50 engineers in 12 months, the lack of a platform layer became a compounding liability: each new squad reinvented deployment patterns, created Snowflake YAML pipelines, and duplicated OCI infrastructure from scratch.
By the time the team reached 30 engineers, deployment frequency had collapsed to twice a week. The company engaged Ksolves to build the internal developer platform that would let them continue scaling without sacrificing the delivery speed that had made them competitive.
Five engineers shipping 10 times a day. Thirty engineers shipping twice a week. The difference was not the people but the absence of a platform.
- Deployment Frequency Fell as Engineering Scaled: As teams grew beyond 30 engineers, inconsistent deployment processes, environment drift, and pipeline conflicts reduced deployment frequency from 10 times per day to just twice a week.
- No Standardized Platform or Golden Path: Every squad independently created OCI deployments, database integrations, functions, and CI/CD pipelines, adding weeks of setup effort before product development could begin.
- Proliferation of Custom YAML Pipelines: The absence of shared CI/CD standards led to dozens of bespoke pipeline configurations that were difficult to maintain, audit, and scale across teams.
- Lack of Service Catalog and Dependency Visibility: With more than 40 services in production, there was no central view of ownership, dependencies, environments, or runbooks, making collaboration and troubleshooting difficult.
- Slow New Engineer Onboarding: New hires required 3-4 weeks of knowledge transfer before making meaningful contributions due to limited documentation and poor service discoverability.
- Inefficient OCI Resource Utilization: Without standardized infrastructure templates or self-service provisioning, teams created duplicate resources and orphaned services, increasing cloud costs and reducing visibility into infrastructure usage.
Ksolves, as an AI-first DevOps consulting services company, built an internal developer platform using Backstage as the central developer portal and Dagger TypeScript SDK as the shared CI/CD engine. The goal was simple: enable any engineer to discover services, understand dependencies, and launch a production-ready OCI application without relying on tribal knowledge or platform teams.
- Standardized CI/CD With Dagger TypeScript SDK: Implemented reusable TypeScript-based pipelines for build, security scanning, OCIR publishing, and deployment to OKE or OCI Functions. Every squad follows the same pipeline structure while retaining flexibility for application-specific logic.
- Backstage Scaffolder Golden Paths: Created production-ready templates for OKE microservices, OCI Functions, and Oracle Autonomous Database integrations. Engineers can provision infrastructure, repositories, and deployment pipelines through a guided self-service workflow in minutes.
- Centralized Service Catalog: Automatically registered every service with ownership, dependencies, environment details, health status, and documentation links, providing a single source of truth for the engineering organization.
- DORA Metrics Visibility: Embedded DORA metrics dashboards within Backstage to track deployment frequency, lead time, change failure rate, and recovery performance across services and squads.
- TechDocs-Powered Documentation: Integrated TechDocs into every template to automatically generate service documentation, architecture records, onboarding guides, and operational runbooks from day one, significantly accelerating engineer onboarding and knowledge sharing.
Technology Stack
| Category | Technology |
|---|---|
| Platform | Backstage + Scaffolder + TechDocs |
| CI/CD | Dagger (TypeScript SDK) |
| Compute | OCI Kubernetes Engine (OKE) |
| Serverless | OCI Functions |
| Database | Oracle Autonomous Database |
| Registry | OCI Container Registry (OCIR) + OCI DevOps |
From twice-weekly deployments and weeks of squad setup to 10 times per day and a production-ready service in under 10 minutes, without asking anyone for help.
- Deployment Frequency Recovering Toward 10x/Day: Standardised Dagger pipelines and Backstage golden paths eliminate environment drift and pipeline conflicts, restoring high-frequency deployments with real-time DORA tracking.
- New Service Setup Reduced to Minutes: Backstage Scaffolder provisions fully production-ready OCI services in under 10 minutes, replacing 2–3 weeks of manual setup.
- Unified CI/CD Across All Squads: Dagger TypeScript pipelines replace fragmented YAML implementations with a single, testable, organisation-wide standard.
- Engineer Onboarding Accelerated to Days: Backstage service catalog and TechDocs enable self-onboarding in under 2 days, replacing weeks of dependency discovery and knowledge transfer.
- Complete Service Visibility Achieved: Central catalog provides real-time ownership, dependencies, and health status across all services, eliminating tribal knowledge and improving cross-team coordination.
“We went from every new hire needing a month to get productive to spinning up their first service on day one – Backstage and Dagger gave us back the velocity we had when there were five of us, but now with fifty.”
– VP of Engineering, Series B Technology Startup
When engineering teams grow faster than their platform, speed doesn’t slow gradually, but it breaks. That’s what happened here: duplicated OCI setups, drifting pipelines, and onboarding that stretched into weeks turned daily deployments into a weekly event. Ksolves rebuilt the system around a unified developer platform. Golden paths replaced reinvention, a single CI/CD standard replaced fragmented pipelines, and a live service catalog replaced tribal knowledge. Now, every service is discoverable, every pipeline is consistent, and every engineer can contribute within days, not weeks. The foundation is built to keep scaling into advanced OCI multi-region delivery, cost governance, and automated compliance as the organization grows.
Is Your Engineering Organisation Scaling Faster Than Your Delivery Platform Can Support?
