Engineers at Lagora use developer tools such as Cloud Code and Cursor.

Lagora uses approximately 10 different AI models, selecting the best one for each task based on latency and performance, with performance being the most important factor.

Jacob Lorentzon hopes to see more European and American open-source models emerge, as he believes a duopoly or monopoly on foundation models would be a negative outcome.

Lagora's developer experience team builds custom tools, including a background coding agent that allows each engineer to run 10 different agents concurrently.

Jacob Lorentzon predicts Lagora's revenue will be above $250 million by the end of the current year.

Jacob Lorentzon believes the future role of engineers will shift from writing code to higher-level systems design, systems architecture, and meta-engineering to make AI agents more effective.

Jacob Lorentzon predicts Lagora's engineering team will grow to approximately 270 people by the end of 2027.

Jacob Lorentzon believes acquiring companies is a faster way to hire groups of A-level talent than individual recruiting.

At Lagora, Product Managers use AI to prototype ideas very quickly, allowing them to test with users before involving engineering.

Lagora has an internal "AI enablement" team that builds internal tools, including custom versions of HR, talent acquisition, and payroll systems, through a process called "vibe coding."

Due to underestimating growth, Lagora now builds all its systems to scale to 100 times their current usage, an increase from a previous standard of 10 times.

Over 50% of the code at Lagora is generated by AI tools, with Claude and Cursor being the top two contributors.

Jacob Lorentzon believes the best available AI model for a given task changes almost bi-weekly, with Lagora switching between models from OpenAI and Anthropic.

Lagora's engineering team has grown to approximately 80 people, significantly exceeding an initial projection of capping out at 20 engineers.

Jacob Lorentzon believes that Cursor could have had a compelling story as an independent, model-agnostic company and was saddened by its acquisition by Groq.

Jacob Lorentzon asserts that AI has made writing code significantly cheaper, shifting the primary bottleneck in software development from coding to code review and product definition.

Jacob Lorentzon predicts that AI code review will become the dominant form of review, removing it as a bottleneck in software development.

Jacob Lorentzon believes the widespread use of AI-generated code creates a new generation of security threats because threat actors can now operate with extreme efficiency.

Lagora uses AI agents to improve the efficiency of its incident response and postmortem processes.

Jacob Lorentzon believes building an engineering organization in 2026 will be very different from 2024 due to AI-driven productivity gains.