According to a recent LinkedIn post from Kilo Code, co-founder Emilie Schario described an engineering model in which company engineers reportedly write about 1% of code directly, with the remainder generated by AI agents. The remarks were made during a panel at VentureBeat Transform 2026 that explored how agentic AI is reshaping software development workflows.
Summer Sale - Claim 70% Off TipRanks
- Unlock powerful investing tools with TipRanks Premium to make smarter, more confident investment decisions
- Subscribe to TipRanks Smart Investor Newsletter, and discover new investing opportunities with data-backed stock picks
The post highlights that the discussion focused on when it may be appropriate to delegate coding tasks to agents versus maintaining a human in the loop, emphasizing safety and oversight considerations. It also notes an emerging cost framework, with “cost per PR” cited as a metric intended to link AI token spend to actual engineering output rather than just tracking overall burn.
For investors, the description suggests Kilo Code may be positioning itself at the forefront of high-automation development practices, potentially enabling greater scalability and margin leverage if quality and reliability are maintained. The emphasis on cost-per-output metrics indicates an effort to manage AI infrastructure spending more rigorously, which could be important for unit economics as AI usage expands.
The focus on agentic AI and workflow redesign also points to Kilo Code’s potential role in the broader transition toward AI-augmented software teams, a theme drawing increasing interest from enterprise buyers and investors. If the company can translate its internal practices and metrics into customer-facing products or services, it could strengthen its competitive standing in the developer tools and AI productivity market.