In this AI startup, the roles have shifted with engineers taking on leadership positions while agents handle the coding tasks. Tatyana Mamut, the CEO and co-founder of Wayfound.ai, an AI agent startup, noted that her team of two engineers is more efficient than larger engineering teams. Mamut, a former director at Amazon Web Services overseeing teams of over 30 people, expressed that her small team can deliver code much quicker. She mentioned that her two engineers outpace the productivity of her Amazon team from 2017.
The transition in Wayfound began in 2024 when engineers started using ChatGPT initially for coding assistance. As new coding tools like Claude Code, Vercel, and Cursor emerged, the team began experimenting with them before settling on Claude Code as their primary tool. Mamut emphasized the remarkable progress made by their engineering team in terms of efficiency due to agents handling testing and code quality monitoring.
Mamut highlighted a shift in how work is managed at Wayfound compared to traditional approaches she experienced in managing large teams. Instead of lengthy processes involving planning, prioritizing, and resolving conflicts, the team now efficiently decides on project requirements during brief meetings on Mondays and Wednesdays. Engineers then execute tasks using Claude Code based on these priorities.
The utilization of coding agents has not only improved efficiency but also led to cost savings for Wayfound by eliminating the need for additional customer success staff. Engineers now engage more with customers rather than spending excessive time in front of screens.
Mamut described the employees’ roles less as traditional engineers but more as “builders,” aligning with a trend observed at companies like Meta where roles across product, design, and engineering are merging into one multifunctional role. She cautioned against the risks of unchecked use of AI tools without ongoing training and supervision by both humans and specialized AI agents.
Regarding the future of software-as-a-service (SaaS) businesses in light of AI advancements, Mamut suggested that companies must evolve into being “agentic” to remain competitive. She emphasized the importance of integrating AI agents into products to stay relevant in the market. Mamut predicted that companies sticking solely to the conventional SaaS model may face challenges in the coming years.
In conclusion, Mamut stressed that while AI agents can handle certain tasks traditionally done by humans, human involvement remains crucial for managing agents effectively, building customer relationships, and setting strategic directions for companies.