Technology

Build the engine of economic ambition at Mandl. Commercial insurance is the quiet force that lets ambition outrun risk, but the outdated and generic system it’s built on is failing our modern economy.

Mandl is an AI company fixing that by rebuilding the engine of commercial insurance from the inside out. In doing so, we’re advancing economic ambition and abundance.

Our roots are in building successful solutions to enormous, unglamorous problems that affect the macroeconomy. Before founding Mandl, our CEO Rushil Goel built Samsara’s (NYSE: IOT) fleet business to more than $500 million in revenue. He then founded Nirvana, a Mandl company, with one of the country’s top actuarial scientists, which has now grown to more than $300 million in run-rate premium with top-tier loss ratios. Vega, our AI-native claims administrator, already resolves claims four times faster than the industry standard. And Cooper, the AI operations platform we built on first principles for the insurance industry, turns 12 to 15 hours of broker work into about ten minutes.

At Mandl, we hire builders who want to create that same kind of impact and scale. If that sounds like you, we’d love to talk. Mandl is backed by General Catalyst, Lightspeed Venture Partners, Valor Equity Partners and others, with $300M+ raised. We’re headquartered in San Francisco.

The Role You’ll build AI agents and AI workflows that enable unique, tailored experiences for insurance use cases. This is a rare chance to own the full arc: design the system, deploy it into production, watch operators rely on it every day, and iterate until it's irreplaceable.

What You’ll Do

• Scale our AI platform: agent orchestration, real-world evals, connectors, data infrastructure

• Deeply understand workflows where someone spends hours, and ship agents that automate them and enable higher impact work

• Build the evals, feedback loops, and training pipelines that teach AI to reliably complete specific jobs, turning capable models into genuine AI coworkers

• Embed with teams onsite, driving adoption beyond the early power users, measuring impact, and shipping the changes that make daily use feel obvious

• Integrate agents with legacy systems with no clean APIs, and making them reliable to run core business operations

Who We're Looking For

• Full stack engineering experience. You’ll be coding across the stack and are comfortable in different languages and platforms.

• Hands-on with modern AI stacks (LLMs, VLMs, ASR / TTS, vector stores, orchestration, evals). You’re pragmatic about trade-offs across latency, cost, reliability and safety.

• Thrive in challenging situations. You’ll map out complex customer workflows and deploy end to end custom builds.

• Strong communication skills. You’ll partner with product and business teams, as well as directly interface with end users.

• High velocity and ownership in an ambiguous environment. We’re moving exceptionally fast in a space that hasn’t been innovated before.

Benefits

• Competitive salary & equity

• Medical, dental & vision insurance (100% employer paid)

• 401k with up to 4% company match

• Unlimited PTO

• Wellness reimbursement

We set our salary ranges using compensation data from companies similar to our stage and size. For this role, the estimated total cash for this position is $180,000- $300,000. The actual total cash salary will vary depending on multiple factors unique to each candidate, including location, job-related knowledge, transferable skills, work experience and our assessment of your level during the interview process. Mandl offers equity for all full-time employees, which is not included in the range listed above and will have an impact on your overall compensation. Mandl is an Equal Opportunity Employer and considers applicants for employment without regard to race, color, religion, sex, orientation, national origin, age, disability, genetics or any other basis forbidden under federal, state, or local law. Mandl considers all qualified applicants in accordance with the San Francisco Fair Chance Ordinance.