How Keller Williams 6X’d Their Deployment Speed with Harness
Initially, Keller Williams ramped up technology development by hiring contractors to rapidly build new systems. While this approach accelerated initial progress, it also resulted in significant technical debt and architectural inefficiencies. The deployment process was cumbersome, requiring manual interventions and consuming excessive time and resources. Prior to adopting Harness, Keller Williams managed only four major software deployments per year, each requiring up to three weeks to complete, followed by additional weeks for post-deployment fixes. The lack of automation and streamlined processes not only slowed down innovation but also negatively impacted the morale of the technology teams.
“When evaluating a new solution, we followed a "rules before tools" approach, defining our ideal feature set before selecting a vendor. Harness stood out due to its ease of implementation, maintainability, and comprehensive feature set that best aligned with our needs.”
- Chris Synan, Vice President of Engineering at Keller Williams
“Since migrating to Harness, we've dramatically increased our deployment frequency. What was once a quarterly process has now become a weekly—and in some cases, daily—occurrence”
- Brandon Taylor-Nelson, Principal Software Architect at Keller Williams
Key Improvements:
Harness provided a cutting-edge CI/CD (Continuous Integration/Continuous Deployment) platform that enabled Keller Williams to automate and streamline its deployment processes. The implementation included:
“With our enhanced CI/CD pipeline, we are delivering updates at the speed of business. Agents and customers now benefit from more frequent improvements, reinforcing confidence in our technology. By ensuring a seamless experience, we help our agents better serve their clients in finding their perfect home.
- Chris Cox, CTO at Keller Williams
Looking ahead, Keller Williams plans to expand their use of Harness by integrating the new Database DevOps product to replace their existing bespoke migration system. Additionally, they aim to enhance their Feature Flags capabilities to facilitate split experiments and improve feature rollouts.