Careers at Nuro: Teams, Pay and How to Get Hired

Who Gets Hired, and Where They Land

The autonomous vehicle sector spent a decade promising fleets. Nuro spent its first eight years building one — then walked away. The pivot from operating delivery bots to licensing a Level 4 stack to automakers and ride-hail platforms didn't just change the business model; it rewrote the org chart. The question now isn't how many vehicles Nuro runs. It's what kind of engineering and operations muscle a company needs when its product is software that drives someone else's car.

Nuro employs about 700 people, Kirsten Korosec's data shows, across nine sites (Mountain View (HQ), San Francisco, Santa Clara, Las Vegas, Houston, Toronto, Washington DC, with Munich, Shenzhen, and Tokyo listed as "Coming Soon") after multiple layoff rounds in 2022 and 2023 and a strategic overhaul that scrapped the company's own low-speed delivery fleet in 2024. The workforce that remains organizes around eight job families published on Nuro's careers page, each mapped to the licensing model: Software & Systems, Hardware & Manufacturing, Operations, Product Management, Commercialization, Technical Program Management, People, and Brand & Marketing.

Software & Systems is the largest family by role count and specificity. Nuro's careers page lists more than 50 distinct titles under this umbrella, spanning the full autonomy stack: Perception (Senior/Staff Software Engineer: Perception ML/Computer Vision; Perception Data Infrastructure; Perception ML Data), Behavior Planning and Prediction (Machine Learning Research Scientist; Senior Software Engineer: Behavior Planning; Senior/Staff Engineer: Behavior ML Data; Technical Lead Manager: Autonomy Evaluation and Intelligence), Localization and Mapping (Senior Software Engineer: Localization and Mapping; Map Platform; Senior/Staff Engineer: Machine Learning Online Mapping), Controls (Senior/Staff Controls Engineer; Senior Software Engineer: Collision Avoidance Testing), Simulation & Eval (Director of Engineering: Eval Platform; Senior/Staff Software Engineer: Behavior Verification; Senior/Staff Machine Learning Engineer: Sensor Simulation; Senior Program Manager: Eval Operations), and the infrastructure layers that make petabyte-scale training deployable on board (ML Infrastructure, Optimization, Data Platform, Labeling Platform, Onboard Infrastructure, Devices Platform, Networking & Real-Time Systems, Sensor Platform, Video Streaming, Performance Tooling). The specificity signals a hiring bar tuned for engineers who have shipped safety-critical autonomy code, not research prototypes.

Hardware & Manufacturing carries two sub-teams, AV Hardware Engineering and NPI Hardware, with roles like AV Hardware Mechanical Design Lead, Senior Electrical Engineer (AV Modules), and NPI Engineer. The scope narrowed after Nuro stopped building its own custom pod; the hardware focus shifted to sensor integration kits and compute platforms that slot into partner vehicles (the Nvidia Drive AGX Thor platform). Operations still runs Central Operations, Fleet Operations, In-Vehicle Operations, and an Operations Executive layer, but the headcount mix flipped: Autonomous Vehicle Operator roles (12-month terms in San Francisco and Houston) and Associate Fleet Technician positions (Bay Area and Houston overnight) now support partner deployments and validation miles rather than a commercial delivery service. Product Management splits between Commercialization Product and Founders' Office; Commercialization lists Staff Product Manager and Head of Product. Commercialization itself sits under Founders' Office, reflecting the direct involvement of co-founders Jiajun Zhu and Dave Ferguson in partnership negotiations: Ferguson told TechCrunch the new capital positions the company to "deliver new commercial partnerships to realize autonomy at global scale." Technical Program Management runs through Software TPM (Senior/Staff Software Technical Program Manager, Systems Engineering). People covers HR, Recruiting (Head of Talent Acquisition, Sr. Technical Sourcer), and Total Rewards.

The geographic distribution of open roles mirrors the testing and partnership footprint: Mountain View and San Francisco concentrate autonomy software, simulation, and product; Santa Clara and Houston host fleet operations and technician roles; Toronto, Washington DC, and the planned Munich/Shenzhen/Tokyo offices map to regulatory, commercial, and OEM engagement tracks. A candidate reading the board sees a company that hires like a Tier 1 autonomy supplier — deep stacks, verification-heavy, partner-facing — not a robotics startup running its own fleet.

What the Pay Looks Like

Nuro's compensation structure reflects its position as a well-funded autonomy company pushing toward commercial scale. The clearest window comes from Levels.fyi's verified employee reports, current as of August 2026, which break out software engineering pay by level. First-party board data from Zero G Talent covers a broader set of roles, including operations, talent, and fleet, but at wider bands.

Software Engineering Levels

The median total compensation for software engineers sits at $300K, with the highest reported package at $574K. Levels.fyi reports a median of $312K. The jump from L4 to L5 is substantial — $139K in total comp — driven largely by stock grants that nearly double. L6 pushes stock to $284K annually, putting senior individual contributors in territory that rivals principal roles at larger tech companies.

Glassdoor paints a lower picture: average $151K, 25th–75th percentile range of $120K–$191K, 90th percentile around $235K, based on 323 salaries submitted. ZipRecruiter reports $141K average. Both aggregate more junior titles and older data; the Levels.fyi figures are newer, level-specific, and sourced from verified offer letters.

Operations, Fleet, and Support Roles

Zero G Talent's board listings (19 salaried roles, band typically $52K–$168K, Zero G Talent's figures put the top there, median $56K) show what the public aggregators miss:

The hourly fleet roles annualize to roughly $52K–$85K depending on overtime and shift differentials. The spread between Bay Area and Houston rates (about 18% at the top end) tracks cost-of-labor differences. Nuro's Houston depot runs 24/7.

Negotiation Reality

Levels.fyi notes most candidates accept the first offer, but their negotiation service averages a $30K improvement. The stock component at L4 and above is the primary lever; base salary bands are tighter. Bonus targets are formulaic and rarely move.

Geographic Context

Nuro's listed sites, the nine locations detailed earlier, suggest compensation will localize. The board data currently covers only Mountain View, Bay Area, and Houston roles. Toronto and DC bands are not yet public. International sites will follow local market rates, not US bands converted at spot FX.

The overall picture: software engineering pays at the top of the AV tier, operations and fleet roles sit at market median for the Bay Area and Texas respectively, and the equity upside at L5/L6 is meaningful if the robotaxi licensing model scales.

How the Hiring Process Works: and What Gets Candidates Through

Nuro's hiring process is designed around a single constraint: the bar is "extremely high, always has been, since day one," said co‑founder and co‑CEO Jiajun Zhu, who still conducts candidate interviews personally and calls it "probably the single most important thing I do." That philosophy shapes every stage, from the first recruiter screen to the final offer review.

Stage 1: Recruiter Screen

Stage 2: Technical Phone Screens

Stage 3: Onsite / Virtual Loop

Stage 4: Hiring Committee & Offer

The process opens with a recruiter call. The recruiter evaluates baseline technical alignment with the role, evidence of the five core values (One Team, Act Now, Earn It, Find a Way, 1% Better Every Day), and whether the candidate's trajectory matches Nuro's scaling phase — growing from one city to multiple cities with an international roadmap over the next two years. Recruiters also flag AI fluency; Zhu has stated the company pushes for "AI fluency across every function, from engineering to operations to how we make decisions," and candidates who already integrate AI tools into their workflow get a stronger signal.

Stage 2 — Technical Phone Screens

Candidates typically face technical phone screens. For software roles these include coding challenges and system design. Robotics and perception tracks add a system‑design component. Interviewers are future peers, and they probe both depth and the "Find a Way" mindset: when the path isn't obvious, does the candidate reason from first principles or reach for a library?

Stage 3 — Onsite / Virtual Loop

The loop runs multiple rounds (ScaleEngineer documents 3–4 rounds for software engineers depending on level, with a difficulty mix of High×3, Hard×2, Med+×2, Med×1 across 62 Glassdoor reports). A typical slate includes deep technical domain questions in the candidate's specialty, system design for a Nuro‑scale subsystem, behavioral and values alignment structured around the five core values, cross‑functional collaboration with a product manager or operations lead, and a conversation with Zhu or a co‑lead. This is not a formality; it's a calibration check on talent density. Candidates often remark they've never spoken to a CEO at this stage elsewhere.

Stage 4 — Hiring Committee & Offer

All interviewers submit written feedback with a hire/no‑hire recommendation tied to specific value scores. A hiring committee reviews the packet. Offers are benchmarked against the first‑party board salary bands and include equity calibrated to the "independent public company" trajectory Zhu outlines.

What Strengthens an Application

• Demonstrated ownership: shipped a safety‑critical feature, led a postmortem that changed process, or built tooling the team still uses.
• AI‑native workflow: candidates who already use AI tools in their work.
• Non‑linear proof: public repos, technical blogs, or competition wins weighted alongside traditional pedigree.
• Prepared questions: interviewers track the questions candidates ask; thoughtful questions signal humility and strategic thinking.

Common Disqualifiers

• Values mismatch: a "no" on core values from any interviewer usually sinks the packet.
• Inability to articulate trade‑offs: candidates who recite textbook answers but can't discuss why they chose one approach over another for a specific operational domain.
• Passive ownership language: "I was part of a team that…" without a clear "I drove X, measured Y, learned Z."
• No AI tooling familiarity: a strong negative signal given the company‑wide AI fluency mandate.
• Over‑indexing on brand names: recruiters have seen candidates from large AV programs who struggle in Nuro's smaller, faster, less‑bureaucratic environment ("team size is a lot smaller… and that's a feature, not a bug").

The process is intentionally thorough because each hire either raises or lowers the talent density Zhu guards. Candidates who treat every round as a mutual evaluation, not a test to pass, tend to clear the bar.

Where the Work Happens

Nuro's physical footprint maps directly to its evolution from a delivery-robotics startup into a technology-licensing platform. As of early 2026, the company operates across its nine sites, the same locations noted above, each serving a distinct slice of the autonomy stack. The distribution follows the hardware, the regulatory milestones, and the partnership agreements that define where Nuro's software meets pavement.

Mountain View: The Brain

The headquarters on the Peninsula anchors the salaried workforce. First-party board data shows four of the five most recent postings tied to Mountain View: Head of Talent Acquisition ($176K–$265K), Total Rewards Manager ($160K–$241K), a 12-month Project Manager for Software Operations ($99K–$149K), and a Software Operations Specialist ($87K–$118K). These roles cluster around people systems, compensation architecture, and the operational software layer that routes, monitors, and supports the fleet. The HQ also houses the leadership team: CEO Jiajun Zhu, CTO Andrew Clare, and the product org that tailors the Nuro Driver for OEM partners like Lucid and mobility platforms like Uber. Clare told The Verge in September 2024 that the licensing pivot was engineered here: "Our tech has gotten to the point where we believe very firmly that it is ready for more applications," he said, noting that eight years ago no OEMs were building L4 platforms, but now "multiple OEMs" are. That shift — from building custom vehicles to integrating the Nuro Driver into nine vehicle platforms — is a Mountain View product decision.

Santa Clara and San Francisco: Fleet Operations

The Bay Area satellite sites handle the physical fleet. Board listings for Associate Fleet Technician roles specify "California - Santa Clara; California - SF" at $30.10–$40.73/hour. These are the depots where Nuro's custom L4 vehicles — half the width of a Toyota Highlander — are staged, charged, inspected, and dispatched. The technician roles are hourly, overnight-eligible, and tied to "Depot 2" nomenclature in Houston, suggesting a standardized depot model. Santa Clara's proximity to the HQ (roughly ten miles) lets engineers iterate on hardware-software integration without shipping vehicles across the country. San Francisco adds urban density: the city's streets were among the first where Nuro logged driverless miles under California's deployment permit, the first such permit issued to any AV operator.

Houston: The Texas Depot

"Texas - Depot 2" appears in the board data for an overnight Associate Fleet Technician role at $25.59–$34.62/hour. Houston has hosted Nuro's commercial delivery operations since the 2018 Kroger partnership launched there, and the 2022 expansion of autonomous operations in Texas made it a proving ground for higher-speed, longer-range runs. A second depot implies fleet scale — multiple vehicles charging and cycling simultaneously — and the overnight shift suggests 20-hour or continuous utilization windows. The pay delta versus the Bay Area reflects local labor markets, not a different job; the technician scope (vehicle prep, sensor cleaning, fault triage, logistics coordination) is consistent across sites.

Las Vegas: Strip Testing at Speed

Nuro began on-road autonomous testing on the Las Vegas Strip in May 2025. The Strip offers a unique stress test: dense pedestrian traffic, complex signal timing, construction zones, and the kind of unpredictable human behavior that highway miles don't replicate. A Las Vegas office supports the safety drivers, remote operators, and data engineers who monitor those runs in real time. The site also positions Nuro for the Uber-Lucid robotaxi program announced July 2025, which targets a 2026 launch and 20,000-vehicle deployment over six years. Vegas is a tourism-heavy, ride-hail-dense market, exactly the environment where a robotaxi service proves its economics.

Tokyo: First International Data Collection

April 2025 marked Nuro's first international data-collection initiative in Japan. A Tokyo office — listed as "Coming Soon" on the company page but actively staffed for the program — handles left-hand traffic, narrow streets, and a regulatory regime distinct from the U.S. The data collected feeds the Nuro Driver's vehicle-agnostic platform, which must generalize across vehicle platforms (nine so far) and geographies. Japan's aging population and dense urban cores align with Nuro's original last-mile thesis; the licensing model makes more sense there than building a local fleet from scratch.

Washington DC and Toronto: Policy and Talent

DC houses the government-affairs and regulatory team that secured the first NHTSA exemption for an AV (2020) and the first California deployment permit. Toronto taps the University of Toronto talent pipeline and gives Nuro a Canadian regulatory sandbox. Neither site shows up in recent board postings, suggesting leaner headcounts focused on specialized functions.

The Expansion Horizon

Munich, Shenzhen, and Tokyo (beyond the data-collection office) signal where the next OEM and mobility partners sit. The "Coming Soon" label means leases signed or offers out, not fully built-out floors.

What the Facilities Enable

Taken together, the sites form a development-to-deployment pipeline: Mountain View defines the product; Santa Clara and San Francisco validate it on Bay Area streets; Houston scales it in a sprawling Sun Belt metro; Las Vegas stress-tests it at commercial velocity; Tokyo proves portability; DC and Toronto manage the rules; the next three cities chase the partnerships that turn the Nuro Driver into revenue. The board's salary band, $52K–$168K median $56K across 19 salaried roles, compresses because most open roles cluster in operations and talent, not senior engineering. That's a snapshot of a company that has already built the core stack and is now staffing the machine that licenses it.

Who Thrives Here

The people who last at Nuro share a recognizable profile: they are drawn to a problem the co-founders themselves describe as "one of the hardest AI challenges of our generation," and they stay because the company's structure rewards a specific blend of technical depth, regulatory patience, and partnership discipline. Jiajun Zhu, co-founder and co-CEO, has said publicly that Nuro exists to "make autonomy accessible to all", not to build a car, not to run a ride-hailing app, not to ship ADAS features. That focus acts as a filter. Candidates who want to own a consumer brand or optimize Level 2 systems self-select out; the ones who remain are comfortable building the substrate that others will integrate into robotaxis, personal vehicles, and commercial fleets.

Leadership coach Darren Gold, brought in to work with the executive team, identified a behavioral marker that distinguishes Nuro's culture: "The strongest leaders don't see tensions as either-or choices. They learn to hold them as both-and possibilities." The tension he references is structural. Nuro must deliver automotive-grade safety, including 1.7 million autonomous miles with zero at-fault incidents, the first NHTSA exemption for a driverless vehicle, and the first California deployment permit, while simultaneously iterating an AI stack that operates across nine vehicle platforms, from the custom R1 and R2 pods to the Lucid Gravity SUVs slated for the Uber robotaxi fleet. It must satisfy regulators in California, Texas, Nevada, and now Japan while moving at the velocity of a software company. People who thrive treat those constraints as design parameters, not obstacles to be complained about.

The pedigree of the founding team sets the technical bar. Dave Ferguson came from Carnegie Mellon's Robotics Institute and the 2007 DARPA Urban Challenge winning vehicle; Zhu was a founding member of Google's self-driving project. They left Waymo in 2016 to pursue commercial delivery, a narrower wedge they believed could reach people faster, offer a sustainable business model, and execute in three to five years. That timeline stretched. The company weathered a 20 percent reduction in November 2022, a 30 percent cut in May 2023 that paused commercial expansion, and a September 2024 pivot to pure technology licensing. The R1 and R2 were purpose-built, zero-occupant delivery vehicles; the Nuro Driver now targets personally owned vehicles and robotaxis on the Uber network. The constant is the autonomy stack, not the chassis.

Regulatory fluency is another differentiator. Ferguson has spoken openly about the work of convincing regulators to certify vehicles that aren't built for humans, as today's rules still require seat belts and airbags. Staff who succeed understand that a software release cadence is meaningless without a corresponding safety case and regulatory strategy.

The partnership model demands a particular collaboration style. Nuro licenses the Nuro Driver and Nuro Toolkit to automakers, fleet operators, and mobility platforms. That means engineering teams must produce interfaces, documentation, and integration support that meet OEM-grade requirements, not just research-grade demos. The Uber-Lucid-Nuro robotaxi program, announced July 2025 and unveiled at CES in January 2026, targets 20,000 vehicles over six years, TechCrunch reported. Delivering on that commitment requires engineers who can translate autonomy capabilities into vehicle-level specifications, validation plans, and production timelines that satisfy both Lucid's manufacturing process and Uber's operational playbook.

A newer internal framework, "autoresearch," signals where the technical culture is heading. Described as "a class of agentic systems where an AI agent runs an iterative experiment loop against a measurable objective and reasons over its own history of attempts," it reflects a shift toward self-improving development pipelines. Engineers who thrive are already comfortable with evaluation-driven development, simulation-at-scale, and closed-loop data flywheels, the tooling that turns 1.7 million miles into the next 10 million.

The vision document frames the ultimate horizon as "Artificial General Robotics", extending the AI infrastructure beyond driving into smart home robots and urban mobility systems. That long arc attracts people who measure impact in decades, not funding rounds. The company's own LinkedIn posts highlight hiring for roles like Head of Product reporting directly to Zhu, with a mandate to "own the product strategy and product roadmap across Nuro as we scale toward widespread commercialization." The language is deliberate: scale, commercialization, roadmap. Not exploration. Not prototype.

In short, Nuro retains people who combine Waymo-level technical rigor with an OEM supplier's delivery discipline, who read regulatory dockets as closely as arXiv papers, and who treat the tension between safety and speed as a design constraint to be managed, not a conflict to be resolved. The layoffs and pivot proved the company will change its business model, its vehicle platforms, and its go-to-market partnerships. It has not changed the problem.

Working in frontier tech? Zero G Talent tracks the openings: see every open Nuro role, browse frontier tech jobs, the companies hiring, and the people building the field.