How Work Gets Done
The Zero G Talent board lists six salaried roles and five internships (controls, embedded systems, mechanical test, shop, and software), all in Berkeley. Salary band: $47k to $111k (Zero G Talent reported), $73k median. One director-level posting: Business Development, $90k (according to Zero G Talent)–$150k (Zero G Talent's board data shows). That org chart is the whole story: a handful of people own the full stack, and the interns aren't fetching coffee. They're writing firmware, machining parts, and closing the loop on test data the same week.
| Role | Pay |
|---|---|
| Director of Business Development | $90k–$150k |
| Salaried median | $73k |
| Interns (all disciplines) | $20–$35/hr |
| Shop Apprentice | $25–$30/hr |
In a shop this lean, the org chart is the dependency graph. The Director of Business Development isn't a layer of management; they translate what the robot does into what a customer pays for. The controls intern sits beside the embedded engineer who sits beside the machinist. Decisions happen at the workbench because there's nowhere else for them to go. A design change that would need three approvals at a 200-person company gets made in a five-minute conversation between the person who wrote the driver and the person who milled the bracket.
No public postmortems, no published handbooks, no founder interviews about process exist. The board data shows a team small enough that "process" mostly means "who's at the bench today." That autonomy is real, but so is the exposure. When the firmware hangs the test stand, the person who wrote it is also the person who has to unhang it, explain it to the mechanical lead, and rewrite the test plan before the end of the day. There's no ticket queue to hide in.
Berkeley puts them in a specific talent current: UC Berkeley graduates, Lawrence Berkeley National Lab alumni, and the dense thicket of robotics spinouts between campus and the marina. The intern pipeline — five distinct technical roles simultaneously — suggests they treat interns as capacity, not charity. At $20–35 an hour, they're paying above academic lab rates but below Series A startup cash. The trade is access to hardware that ships.
What isn't visible in the data: how priorities get set when the founder and the lead engineer disagree, whether there's a weekly rhythm or just a Slack channel that never sleeps, and how they handle the weeks when the robot works and the weeks when it doesn't. The board shows the shape. The texture is still private.
What the Board Reveals About Culture and Hiring
The public record describes three unrelated entities named Terranova: a muralist running Colorful Matt Design, Fortra's cybersecurity awareness division, and a summer events program in Calabria. None reference the Berkeley robotics company posting for controls, embedded, mechanical test, shop, and software interns on the Zero G Talent board.
That absence is the signal. A lean team (six salaried roles, a $47k–$111k band, a single director hire) doesn't publish a polished values page. The board structure suggests a founding team stretching capital across technical hires while keeping cash compensation modest. In early-stage hardware, that pattern correlates with an unwritten operating principle: scope ownership over role definition. The intern titles map to the full stack of a physical product. A Shop Apprentice at $25–$30/hr working beside a Controls Intern at $25–$35/hr implies a shop floor where hierarchy is flat because the work demands it.
The posting set reveals a preference for breadth and hardware fluency. The company isn't hiring a "frontend engineer" or a "DevOps specialist." It's hiring interns who can cross mechanical, electrical, and firmware boundaries. That hiring shape reflects a founder priority: people who unblock themselves across disciplines without waiting for a handoff. The Shop Apprentice role, rare in pure software shops, signals that physical fabrication is a daily activity, not an outsourced step. In that environment, the operating principle is effectively "if it needs building, you build it," and the cultural filter selects for candidates who treat a mill, a soldering iron, and a debugger as equally familiar tools.
A team of six salaried roles cannot support specialized managers for each function. Decision-making authority sits with the people doing the work because there is no one else to escalate to. That is not a stated value; it is a structural necessity. The research contains no founder quote articulating "move fast, own the outcome," but the org chart writes it in budget lines.
Fortra's Terranova Security press release quotes executives on reducing "user-initiated cyber risk." The UD profile of Matt Terranova quotes him on wanting "your eye to move from vignette to vignette." Neither belongs to this company.
If you are evaluating this Terranova, treat the job board as the clearest cultural artifact available. The roles posted, the bands offered, and the Berkeley location (hardware talent density, prototype shop access) collectively describe a company that values demonstrated cross-domain output over credentialed specialization. The operating principle you will encounter daily is likely unwritten: the person closest to the problem decides the fix. That works for engineers who have shipped hardware from concept to test; it frustrates those who expect a spec, a manager, and a handoff. In early-stage robotics, the hiring plan is the culture doc.
The HireVue research in the digests describes a generic AI screening layer used by many large employers: three to six video questions, two to three game questions, 15 to 25 minutes total. Terranova does not appear in that research. If they use an asynchronous screen, it would be a filter, not the decision. The real assessment happens when a candidate sits across from founders who need to know: can this person operate without a spec, debug a test stand failure at 11 p.m., and ship a fixture design by Friday because the integration window moved up?
The intern slate reveals the technical floor. Controls, embedded, mechanical test, and software are not adjacent disciplines — they are the full stack of a robotic system. A shop apprentice role signals that hands-on fabrication and fixture building are treated as engineering work, not technician work. Candidates who have only simulated in Gazebo or written ROS nodes on a clean Ubuntu VM will struggle. The bar selects for people who have touched hardware, broken it, and fixed it without a lab manager handing them the next tool.
For the Director role, the $150k ceiling in Berkeley sits below market for a VP-level BD lead at a Series A startup. That gap is intentional. The hire Terranova can afford — and the hire who will stay — is someone who values ownership over title, who has carried a quota in a technical sale, and who can translate a robot's capability into a customer's ROI without a marketing team writing the deck. A mis-hire at this scale doesn't just cost recruiting fees; it stalls the one program the founder isn't personally running.
The hiring bar, in practice, selects for three signals the board data implies: demonstrated breadth across the hardware/software boundary, evidence of shipping with minimal supervision, and a tolerance for ambiguity that reads as excitement rather than anxiety. No formal "culture add" rubric or structured interview panel is documented. In a six-person salaried team, the panel is the company. If the founders don't see themselves trusting you with a subsystem by week three, the process ends — no matter how well you scored on a generic competency model.
What Employees Say (Nobody's Talking)
That job board shows six salaried roles and a handful of hourly internships, but those postings describe compensation and role scope, not lived experience.
The research corpus supplied for this article contains zero employee testimonials, internal survey results, or attributed quotes from current or former staff. What appears instead are two unrelated bodies of material: a 2018 academic event featuring sociologist Tiziana Terranova discussing platform capitalism, and a 2023 Florida real-estate press release announcing that Terranova Corporation (a Miami Beach commercial landlord) leased 2,751 square feet to a candy retailer. Neither references the Berkeley operation.
This silence is itself a data point. Early-stage hardware companies with small teams often fly under the review-site radar simply because headcount is too low to generate statistically meaningful samples. Candidates should treat the silence as a prompt for direct due diligence: ask hiring managers for references to recent interns or full-time hires, request a brief conversation with a current engineer on the controls or embedded team, and probe how the founder-led structure translates into day-to-day feedback loops, performance review cadence, and equity transparency. The job postings confirm a multidisciplinary hardware program (controls, embedded, mechanical test, shop apprenticeship, software) suggesting a lab-to-floor culture where iteration cycles are tight and cross-functional ownership is expected. But without attested employee voices, the only reliable way to gauge whether that pace feels like autonomy or burnout is to hear it from someone inside the room.
Who Thrives, Who Burns Out
The board data shows six salaried roles and a stack of intern openings, a ratio that signals a lean core team building out technical depth fast. In that environment, the people who last treat ambiguity as a design parameter, not a blocker. They show up on day one knowing how to move a hardware project from a whiteboard sketch to a test article without a program manager handing them a spec. They write their own test plans, debug their own firmware, and walk to the shop floor when a tolerance stack doesn't close. The Director of Business Development role suggests the founders expect that same ownership on the commercial side: one person who can run a sales cycle, negotiate a contract, and translate customer requirements back to engineering without a handoff chain.
People who need a clear lane struggle. If your instinct when a requirement shifts is to ask for an updated Jira ticket, you will wait a long time. The intern pay bands reflect a structure where junior hires are expected to contribute real work immediately, not shadow a mentor for a quarter. The shop apprentice sits beside the embedded systems intern; both are close to the hardware, both are expected to close loops. Candidates who have only operated in large orgs with dedicated test labs, procurement teams, and configuration management often find the pace disorienting. They spend their first month looking for the process that doesn't exist yet.
The median salaried compensation of $73k sits below what a senior engineer commands at a prime contractor or a well-funded Series B. That gap is intentional. The trade-off is scope: a controls engineer here owns the control loop, the actuator selection, the safety case, and the demo that wins the next contract. At a larger shop, those split across three roles. The person who thrives treats that breadth as the reward. The person who burns out treats it as scope creep.
Founder-led means the founders are in the lab, on the customer calls, and in the hiring loops. They set the tempo by example. If you need a manager to prioritize your week, you will not get one. You also won't get a manager to shield you from the customer. The ones who stay are the ones who would rather argue directly with a founder about a technical direction than wait for a consensus meeting that never happens. They document their decisions because no one else will, and they ship because the runway depends on it.
The intern-heavy pipeline also means mentorship is informal and uneven. A software intern may pair with a founder who hasn't written production code in two years, or with a senior engineer who is also the only person who knows the build system. Candidates who have only learned in structured programs (bootcamps, co-ops with defined rotations, graduate research with a PI who meets weekly) often hit a wall when the only code review is "ship it and we'll fix it in the field." The ones who adapt fast are the ones who have already built something end-to-end on their own: a CubeSat bus, a robot arm, a custom PCB that flew on a drone. They know what "done" looks like without a rubric.
Burnout tends to arrive not from hours — though hours happen — but from the cognitive load of holding the whole system in your head while the requirements move underneath you. The mechanical test engineer who also has to fixture the part, write the LabVIEW script, and explain the failure mode to the customer on a Friday afternoon is doing three jobs. The person who thrives sees that as the job. The person who leaves sees it as a failure of organization. Both assessments are accurate. The company is not going to hire the three specialists to relieve the load until the revenue justifies it. That is the bet every hire makes — the same bet the board data showed in the first paragraph: six people, full stack, no safety net.
Working in robotics? Zero G Talent tracks the openings: see every open Terranova role, browse robotics jobs, the companies hiring, and the people building the field.