Agility Shipped 150 Robots to Paying Customers. Tesla Shipped 150 to Itself. The Difference Is a Factory in Oregon Hiring at $186,000.

The RoboFab Signal

Agility Robotics' careers page lists over 60 open roles. A growing share point to one location: Salem, Oregon, where the company's RoboFab factory is shifting from construction into live production and hiring across manufacturing engineering, assembly, and NPI.

The signal is in the job titles. Agility is hiring a Senior Manager, Manufacturing Engineering, NPI for Salem, a role focused on "bridging the gap between design and production" and leading "development and validation of manufacturing processes for new products, from prototype to full-scale production." That language is specific. This is not an R&D lab role. It is a production-ramp role, and it sits at the center of what the company is building in Oregon.

The same pattern shows up further down the org chart. Agility is also hiring a Senior Assembly Technician I for the RoboFab, responsible for "assembly, inspection, test, repair and refurbishment" of humanoid robots at the Salem factory. The company's careers page adds a Staff NPI Manufacturing Engineer and a Senior Service Technician II, both on-site in Salem. Zero G Talent's board data shows five Agility roles added in the past week alone, with three of the five tied to Salem.

Taken together, the hiring profile tells a clear story: the RoboFab is moving past the build-out phase and into the operational phase. Agility isn't just designing humanoids in Salem anymore. It is staffing up to manufacture, test, and service them there at scale, a transition that most humanoid-robot companies have not yet reached.

Two Deployments in Ten Weeks Changed Everything

On December 10, 2025, Agility announced a commercial agreement with Mercado Libre to deploy Digit humanoid robots at the Latin American e-commerce giant's fulfillment facility in San Antonio, Texas. The partnership targets "high-turnover, hard-to-staff roles" (repetitive, physically taxing tasks in commerce fulfillment) with plans to expand across Mercado Libre's Latin American warehouse network. Agility's chief business officer Daniel Diez called Mercado Libre "a true innovator in both commerce and fintech." Mercado Libre's senior VP of shipping Agustin Costa said the deployment was "a significant step forward in our vision to create a safer, more efficient, and adaptable logistics network."

Then on February 19, 2026, Agility announced a second deal: Toyota Motor Manufacturing Canada signed a commercial agreement to deploy Digit at its Woodstock, Ontario facility after a completed pilot. The scope here is broader (manufacturing, supply chain, and logistics operations), and the customer is one of the world's largest automakers.

Two customers. Two countries. Two very different operational environments. And both require the same thing from Agility's Salem workforce: people who can get robots out of the lab and onto the floor.

The jobs board tells the story. Zero G Talent's board shows Agility added five roles in a single week, spanning the exact functions these deployments demand:

These aren't research positions. The NPI engineer bridges prototype and production: the person who figures out how to build Digit reliably at volume once the design is locked. The integration engineer shows up at a Mercado Libre warehouse in San Antonio or a Toyota plant in Ontario and makes sure Digit actually talks to the customer's existing systems. The service technician keeps the robots running after deployment, handling on-site maintenance and troubleshooting.

Mercado Libre's deal is notable for its geographic ambition. The San Antonio facility is the starting point, but the companies explicitly plan to "explore additional use cases where AI-powered humanoids can augment logistics operations across Mercado Libre's warehouses in Latin America." That means Agility needs field-deployment engineers who can work in new facilities, adapt Digit to unfamiliar workflows, and train local staff, repeatedly, across multiple sites.

Toyota Canada's deal adds a different pressure. Automotive manufacturing environments have tighter safety requirements, more complex material flows, and higher stakes for downtime. TMMC's deployment covers "manufacturing, supply chain and logistics operations," which means Digit has to perform in settings that look nothing like a warehouse aisle. The integration work is harder. The support burden is heavier.

Both deals also lean on Agility's Arc platform, the cloud-based system for fleet management and enterprise integration. Arc lets Digit coordinate with other automation and slot into existing operations without facility redesigns. But someone has to configure, maintain, and troubleshoot that integration at each new site. That's the hiring surge hiding behind the press releases.

Mercado Libre joins GXO, Schaeffler, and Amazon as companies deploying Digit. Toyota Motor Manufacturing Canada adds a major automotive name to the list. Each new customer doesn't just mean more robots shipped; it means more integration engineers dispatched, more service technicians on call, more production staff in Salem building units to keep pace. The commercial agreements are the visible part. The job postings are what the commitments actually cost in headcount.

What Scaling Humanoids Actually Requires

The company's site lists over 60 open positions across its three US offices, and the breakdown maps the gap between a working prototype and a product that ships. It's not just AI researchers writing motion-planning algorithms. It's manufacturing engineers, supply chain managers, field technicians, and functional-safety specialists.

The NPI backbone

The most telling cluster sits in Agility's Operations division. The company is hiring a Staff NPI Manufacturing Engineer and a Staff Technical Program Manager, NPI, both based in Salem.

These aren't research positions. The NPI Manufacturing Engineer owns the transition from prototype to production line, designing workflows, qualifying tooling, and making sure a robot that works in a lab can be built repeatedly in a factory. The Technical Program Manager coordinates across design engineering, supply chain, and operations to keep that transition on schedule. Both roles exist because Agility is no longer proving Digit can walk. It's proving Digit can be manufactured.

A Senior Manager, Manufacturing Engineering, NPI role caps this track. The job description calls for 10 years of manufacturing engineering experience, five in leadership, with expertise in DFM, lean principles, and Six Sigma. This is a role borrowed from automotive and electronics manufacturing, now applied to bipedal robots.

Hardware depth across Fremont and Pittsburgh

Agility's Fremont, CA office, which the company describes as "where the AI revolution is happening," is the hub for hardware engineering. Open roles there include a Staff Firmware Engineer, a Staff Systems Validation Engineer, a Staff Hardware Test Engineer - Torso, and a Staff Systems Engineer, Functional Safety. The specificity of these titles matters. "Hardware Test Engineer - Torso" means Agility has moved past testing the robot as a single system and is now validating individual subsystems at scale. "Functional Safety" is a role that exists in automotive and aerospace, industries where a failure kills someone. Its presence on a humanoid-robot job board signals that Agility is building for deployment environments where safety certification is a requirement, not an afterthought.

Pittsburgh handles "engineering testing, validation, and skills development." The Engineering Support Technician role there is on-site, hands-on work: the person who keeps test rigs running and robots assembled during validation cycles.

Field deployment: the new category

The roles that didn't exist in robotics five years ago are the ones tied directly to Agility's commercial deals. A Solutions Engineer II position listed on LinkedIn calls for 2–5 years in robotics deployment, field engineering, or systems integration, with proficiency in Linux, Python, and hardware/software APIs. This is the person who shows up at a Mercado Libre warehouse or a Toyota Canada facility and makes Digit work in the real world, not in a controlled demo, but alongside actual workers doing actual tasks.

The Senior Service Technician II role in Salem is the counterpart: the person who maintains deployed robots, diagnoses failures in the field, and feeds reliability data back to engineering. Zero G Talent's board lists this role among Agility's five most recent postings.

Software and AI: still core, now specialized

Agility's Software Engineering division remains the largest single category, but the roles have narrowed in scope. Instead of generic "robotics engineer" titles, the listings specify Senior Software Engineer, Autonomy Core, Staff AI Engineer, Perception, Staff Robotics Software Engineer, Robot Behaviors, and Staff Software Engineer, Teleoperation. Each maps to a specific layer of the autonomy stack. The company isn't hiring generalists to explore what's possible. It's hiring specialists to harden what already works.

The supply chain layer

Agility is also hiring a Global Supply Manager - Ecosystem Devices, a Sr. Supplier Quality Engineer, and a Buyer II. These procurement and quality-assurance roles determine whether a robot can be built from parts that actually arrive on time and meet spec. The Supplier Quality Engineer spans both Fremont and Salem, tying component qualification directly to the production floor.

Strip away the job titles and a pattern emerges. Agility is hiring for three distinct phases of the product lifecycle in parallel: design and validation (Fremont/Pittsburgh), manufacturing scale-up (Salem), and field deployment and support (Salem plus remote). Most robotics companies at this stage are still concentrated in the first phase. Agility's hiring spread suggests it's trying to build all three at once, which is either a sign of serious commercial momentum or a company stretching itself thin across the hardest parts of the robotics value chain simultaneously.

Why Salem Matters More Than Silicon Valley Right Now

The humanoid robot sector is crowded with well-funded players, but most are still in the lab. Tesla's Optimus program, despite Elon Musk's projection that it will be "worth more than the car business," had shipped roughly 150 units as of early 2026, all for internal testing inside Tesla factories, with none performing productive tasks on the line. Figure AI, which has raised $1.9 billion and carries a $39 billion valuation, shipped a similar number in limited enterprise pilots. Both companies remain in what the industry calls "pilot" status.

Agility shipped about 150 units too. But those went to Amazon warehouses and, more recently, to Mercado Libre and Toyota Canada, commercial deployments with paying customers. That distinction changes what the company needs to hire for. Tesla and Figure are still largely hiring for R&D. Agility is hiring to build, ship, and support robots that are already working.

The Salem RoboFab is the physical proof of that shift. At 70,000 square feet, it is the world's first full-scale factory purpose-built for humanoid robot manufacturing, with a peak capacity of 10,000 units per year. Agility designed it around modular workcells that can expand without heavy industrial machinery, a deliberate choice that lets the company scale production without the capital overhead of a traditional auto plant. The facility opened in late 2023 and is now ramping toward volume output.

Compare that to the competition's manufacturing posture. Tesla plans to mass-produce Optimus in 2026 and has targeted a price point of $20,000–$30,000 per unit, but analysts widely expect those targets to slip. Figure AI's Sunnyvale headquarters houses 800-plus employees, and the company is hiring aggressively. Zero G Talent's board lists five Figure roles added in the past week alone, including a Helix AI Engineer for Perception at $200,000–$350,000 and a Mechanical Engineer for compliant hand elements. But Figure's manufacturing strategy still relies on external partners rather than a dedicated factory.

Agility's Salem operation is pulling a different kind of talent. The five most recent roles on Zero G Talent's board include a Staff NPI Manufacturing Engineer in Salem and a Senior Service Technician II, also on-site in Salem. These are not research positions. They are the people who turn a working prototype into a product that ships, survives a warehouse floor, and gets maintained by a technician who may not have a robotics PhD.

The humanoid industry's bottleneck is no longer AI. Figure's Helix VLA model runs at 200Hz, Tesla's FSD chip processes visual data at 144 trillion operations per second, and both companies have solved the "walking" problem. The bottleneck is manufacturing: building robots reliably, affordably, and in volume, then keeping them running in the field. Agility's RoboFab is the only facility in the sector designed to do exactly that at scale.

For engineers deciding where to work, the signal is straightforward. Tesla and Figure are building the brain. Agility is building the body, and the factory, and the supply chain, and the service organization around it. Salem is where the industry's first real production jobs in humanoid robotics are appearing, and the Mercado Libre and Toyota Canada deals are the reason why.

What the Hiring Signals Mean for Robotics Engineers

If you're an engineer or technician weighing a move into humanoid robotics, the window is open, but where you land depends heavily on choices you make now.

Salary benchmarks: Oregon vs. the national picture

Robotics engineers in Oregon earn a median around $139,000 per year, according to Salary.com, with a range from $117,000 to $157,000. Glassdoor's Portland-area estimate puts total compensation slightly higher at $130,000, including roughly $27,000 in additional pay. Those figures sit below the national robotics median of $156,563 reported across 907 jobs with disclosed salaries, but Oregon's lower cost of living narrows the gap in real terms.

The more important number is the range within the field. A 2025 analysis of robotics salaries found that software-focused roles (machine learning engineers, motion planning engineers, robotics software engineers) cluster between $189,000 and $205,000. Hardware-focused roles (automation technicians, field service engineers, electrical and mechanical engineers) average $127,000. That 53% premium for software tracks compounds over a career. The difference between a $194,000 average and a $127,000 average exceeds $2 million over 30 years.

Agility's own postings reflect this spread. These map directly to the national data.

What skills actually move the needle

The highest-paying individual skills in robotics job postings are CUDA (+45% premium), technical leadership (+43%), and machine learning (+40%). But the data cautions against reading those as individual skill premiums. They're proxies for career tracks. CUDA appears in just 6% of postings, so it signals a specialized software role rather than a resume line that bumps your offer by 45%.

The practical takeaway: Python shows up in 43% of robotics jobs, C++ in 38%, and machine learning in 31%. Those three form the baseline toolkit for the software track. If you're coming from a hardware or controls background and want to switch, you need the full stack, not a single course. Employers hiring software roles expect Python, C++ or both, an ML framework like PyTorch or TensorFlow, and domain depth in something like computer vision or motion planning.

For engineers staying on the hardware side, the picture isn't bleak. Controls engineers in aerospace and defense earn a median of $184,000, well above the $109,000 average for that role in industrial manufacturing. Industry choice matters as much as role choice.

The experience curve rewards early aggression

Salary growth in robotics is front-loaded. The jump from entry-level ($89,000) to junior ($115,000) delivers 29% growth. Junior to mid-level adds 35%, the steepest climb in the career. Mid to senior still delivers 25%. After that, the curve flattens: senior to lead adds only 11%, roughly $20,000.

The first five years are where salary acceleration happens. Engineers who stack high-impact projects and build specialized skills during that window capture most of their lifetime earning potential. After reaching senior level, optimizing for base salary alone makes less sense. Equity, technical depth, and work-life balance become the real differentiators.

Practical takeaways

If you're early in your career and targeting humanoid robotics, the data points in two directions. First, get on the software track if you can: the premium is large and persistent. Second, don't over-invest in advanced degrees for salary ROI. Master's degree holders earn a median of $186,000 versus $169,000 for bachelor's holders, a 10% bump that takes decades to recoup in lost income and tuition. A PhD actually pays slightly less than a master's in this market ($180,000).

If you're mid-career and already in robotics, your single biggest lever is industry switching. A "Research Scientist" in academia earns $112,000. The same title in transportation and autonomous vehicles pays $200,000. No new skills required, just a different employer. Engineers in industrial manufacturing earning around $102,000 can roughly double their income by moving to a robotics software or autonomous vehicle company.

For technicians and field-service engineers, the path runs through specialization and employer choice. The gap between a $69,000 automation technician role and a $183,000 senior hardware engineer role represents 167% growth potential, but it requires deliberate skill-building over years, not a lateral move.

Agility's Salem factory is hiring across all of these tiers right now. The specific roles (NPI engineers, systems integration, field technicians) map to the national salary data closely enough that you can benchmark any offer against these figures with confidence.

Working in robotics? Zero G Talent tracks the openings: browse robotics jobs, openings at Figure AI and Agility Robotics, and the people building the field.