Why We're Building an AI Tech Lead

Why Now

Modern engineering teams face a fundamental tension: the need to ship faster while maintaining technical excellence. Growth means more developers, more repositories, more moving parts, but human leadership doesn’t scale at the same rate.

At Rippling, we lived this problem every day. With hundreds of micro-services and globally distributed teams, even the most experienced engineering leaders struggled to maintain consistent code quality, enforce architecture, and mentor effectively at scale.

Traditional approaches (eg, senior engineers manually reviewing every pull request) tribal knowledge passed over Slack—simply don’t work as organizations grow. And while AI tools have started helping individual developers code faster, they aren’t addressing the real bottleneck: engineering leadership.

The opportunity is clear. Advances in AI, large language models, and system graphing make it possible to build something new, not just a code assistant, but an AI Tech Lead: a system that can reason across your codebase, architecture, and goals, and provide the leadership needed to build great software at scale.

We believe the teams that win in the next decade will be those that successfully combine human creativity with AI judgment, and it’s time engineering leadership caught up.

Where We’re Starting

The most natural place to begin is code review, the daily heartbeat of software quality.

But we’re approaching it differently.

Propel doesn’t just look at a diff. It understands your codebase architecture, the patterns you’ve chosen, the modules you rely on, and the downstream impact of every change. It reviews pull requests like an experienced tech lead would: not only catching bugs, but flagging architectural drift, suggesting better design patterns, and highlighting risks you might not see immediately.

In early pilots, Propel is already catching performance regressions, security gaps, and architectural mismatches that lightweight AI reviewers miss. It’s thoughtful, context aware, and focused on helping your team improve, not just approve.

We see AI code review not as the final product, but as the first proof point. The first place an AI Tech Lead can immediately start adding value to real engineering teams.

Introducing Propel

Propel is the AI Tech Lead built to work alongside your engineering team, reviewing, mentoring, and guiding with full system awareness.

With Propel, teams can:

• Get real, context rich code reviews in minutes, not days

• Catch architectural inconsistencies early, before they snowball

• Mentor developers automatically through actionable, system aware feedback

• Prioritize risks intelligently based on system impact, not just code style

• Align technical decisions with product and business outcomes

Unlike lightweight AI tools that treat code as isolated snippets, Propel understands how your systems work together. It knows when a small change in one service could ripple across your platform. It knows when a shortcut in a module could turn into tech debt six months from now.

And most importantly, it’s built to raise the bar, not lower it.

Propel isn’t about getting code merged faster at any cost. It’s about helping your team merge better code, faster.

We’re building an AI teammate that earns trust the way great engineers do: through judgment, helpfulness, and raising the standard without slowing the team down.

Where We’re Going

Our vision is bigger than code review. We believe the next evolution of AI in engineering isn’t just making developers faster—it’s making entire teams better.

In the future, Propel will:

• Review architectural decisions and design docs with system level reasoning

• Analyze and prioritize technical risks before they become outages

• Onboard new developers by answering codebase and system questions instantly

• Map technical changes to KPIs and product outcomes, helping leaders steer better

• Enforce security, compliance, and engineering standards continuously and automatically

In other words, everything a great tech lead does, at scale.

We know this won’t be easy. Building an AI that can reason across systems, understand evolving architectures, and adapt to each company’s unique style and standards will take real engineering rigor.

AI today is powerful, but imperfect. The frontier is crowded. Progress will be hard earned.

But we believe it’s possible because we’ve felt the need for it ourselves. We know the impact it would have had on our teams. And we know that if we build it right, Propel won’t just save teams time but will make them meaningfully better.