How Work Gets Done
Heron Power is building a 40-gigawatt transformer factory before it has a backlog to fill — a bet that the grid's hardest bottleneck is physical, not financial. Drew Baglino, who oversaw Supercharger deployment and utility-scale battery projects at Tesla, founded the company after concluding that transformer lead times, not capital, are the binding constraint on decarbonization. "While there was a lot of innovation happening as far as the technologies that connect to the grid, the actual stuff on the other side of the wire that you're getting from utilities or that you're buying from suppliers like a transformer isn't changing at all," he told Forbes. "And if it's really painful to expand the electricity sector because of bottlenecks in the supply chain, something should be done about that. That's why I started the company."
That frustration orients every engineering and manufacturing decision at the Scotts Valley site. The research reveals a coherent operating system: one that tackles the toughest physical constraint first, prices for zero-friction adoption, hires only hard-tech veterans, and builds capacity ahead of demand — but it does not reveal daily workflows. No founder interviews, internal communications, employee reviews, or organizational charts describe sprint cadences, design review processes, prototype iteration loops, or how engineering decisions escalate. The salary data from Zero G Talent's board (median $195k across 17 salaried roles) implies a hardware-focused organization hiring across power electronics, firmware, automation, and business applications. But without grounded evidence, any description of "how work actually gets done" at the team level would be fabrication. The gap is total on the ground; the signal is clear only at the system level.
Values and Operating Principles
Baglino's frustration translates into a product philosophy that prizes density and speed over feature sprawl. Heron's solid-state transformers are roughly 30 percent smaller and weigh a quarter as much as conventional units, yet they transfer more electricity with lower losses and install faster. The company prices them at parity with traditional transformers, which can exceed $200,000 for large data-center models, betting that reduced installation time and operational waste will drive adoption. That pricing discipline reflects a second principle: the new hardware must slot into existing procurement and interconnection workflows without demanding a premium. Customers named so far include developers of utility-scale solar and battery projects and hyperscale data-center operators such as Intersect, a signal that the initial go-to-market motion targets buyers who measure downtime in revenue per minute.
A third principle is visible in the capital structure and advisory bench. The $140 million funding round, with Andreessen Horowitz and Breakthrough Energy Ventures among the investors, followed a $43 million raise a year earlier — capital sized for factory construction, not software iteration. Advisers JB Straubel (Tesla co-founder) and Zachary Kirkhorn (former Tesla CFO) bring direct experience scaling hardware from pilot line to gigawatt-hour volumes. Their presence signals that Heron's operating model mirrors the Tesla playbook: vertical integration of power electronics, aggressive automation targets, and a tolerance for upfront capital intensity that pure-play software investors typically avoid. The first automated factory, location still undisclosed, is slated to come online before the end of 2027 with a design capacity of 40 gigawatts per year — roughly half of Texas's annual generation.
That factory target reveals a fourth principle: throughput as a proxy for grid impact. By building for 40 GW of annual output, Heron is sizing its manufacturing to the bottleneck, not to a near-term order book. The company's public communications frame this as "reducing waste in the process" of moving renewable energy onto the grid — a phrasing that doubles as an efficiency metric and a cultural cue. Engineers at Scotts Valley are not optimizing for lab benchmarks; they are optimizing for a production line that must yield shippable units at a rate that makes a measurable dent in interconnection queues.
None of these principles appear in a framed values poster. They appear in the founder's public quotes, the spec sheet the company publishes, the investors it chose, the advisers it recruited, and the factory capacity it committed to. Taken together they form a coherent operating system: solve the hardest physical bottleneck first, price for frictionless adoption, staff with people who have already scaled hard tech, and build the factory before the backlog forces you to.
What the Hiring Bar Selects For
The research on Heron Power's specific hiring criteria is thin. The primary source material (a Bloomberg Tech interview with the founders of Byteboard, a technical interviewing platform) articulates a philosophy that criticizes industry norms rather than documenting Heron's own process. Byteboard's founders said traditional software engineering interviews "focus on what you learn in a computer science classroom so highly theoretical knowledge" and are "completely broken and disseparate from the way that engineers work on the job day to day." They cited data claiming only a third of candidates perform consistently across interviews, and that women are seven times more likely to drop out after a poor interview experience. "It's not candidates that are failing the technical interview process it's the technical interview process that is failing them," they argued. Their counter-move is a project-based assessment that "actually evaluates for skills that are used on the job," explicitly tied to pipeline diversity: "we have a leaky pipeline that is just validating you know the perspective of you know when i look at the industry i don't see people that typically look like me."
Whether Heron adopts Byteboard's tool or its philosophy is not documented. What the research does establish is Heron's live job board: six open roles in Scotts Valley as of this writing, spanning firmware, power electronics, automation, product, and sales, with salary bands ranging from $101,000 to $228,000, and a median of $195,000 across 17 salaried listings.
| Role | Salary Band |
|---|---|
| Head of Sales | $200k–$240k |
| Staff Firmware Integration Engineer | $170k–$240k |
| Automation & Controls Engineering Manager | $170k–$220k |
| Staff Product Manager, Business Applications | $180k–$200k |
| Electronics Design Engineer | $150k–$200k |
| Senior/Staff Power Electronics Engineer | $120k–$200k |
Median across 17 salaried listings: $195k
The profile suggests a need for senior technical contributors who can integrate across hardware, firmware, and controls, aligning with Byteboard's stated emphasis on job-relevant, project-based evaluation over algorithmic recitation. But without Heron's own interview guides, debrief notes, or hiring manager commentary, any claim about what the company's bar actually selects for would be extrapolation. The research gap is real: a coherent critique of broken industry defaults and a product built to fix them, not a portrait of Heron Power's hiring bar.
What Current and Former Employees Say
Public feedback on Heron Power is fragmented across two Glassdoor profiles that may or may not represent the same operating entity — a detail that complicates any clean summary. The "Heron Agency" profile carries a 3.7 out of 5 rating across 26 reviews, a sample size large enough to suggest a consistent signal rather than a handful of outliers. That aggregate sits in the "okay to good" band typical of early-stage hardware companies where mission alignment offsets operational chaos.
A separate Glassdoor listing under simply "Heron" shows only five reviews total, with three visible in the preview. The verbatim pros and cons read like shorthand: "Was good not much issues" on the positive side; "Not good pay hours too little" on the negative. The brevity of both entries makes it impossible to tell whether reviewers are describing engineering roles, sales, operations, or contract work. The platform does not date-stamp individual reviews in the snippet view, so the timeline of sentiment is opaque.
The entity ambiguity matters. Heron Power's Scotts Valley roles carry salary bands well above the "not good pay" complaint on the generic Heron profile, but the Glassdoor reviews could predate the current compensation structure, reflect non-exempt roles, or belong to a different Heron altogether. Without review dates or role tags, the conflict cannot be resolved — only flagged.
What the data does support is a pattern common to capital-intensive startups: a modest overall rating that masks wide variance by function and tenure. Hardware teams at this stage often rate culture highly on technical autonomy and mission clarity while scoring compensation, work-life balance, and process maturity lower. The Heron Agency aggregate of 3.7 aligns with that template. The five-review Heron profile is too thin to generalize from, but its lone con (pay and hours) echoes the most frequent complaint in early-stage power-electronics shops where tape-out deadlines compress schedules.
No first-hand employee narratives, exit interviews, or longitudinal sentiment trends are available in the research. The Zero G Talent board data confirms Heron Power is actively hiring across senior technical and commercial tracks in Scotts Valley as of the current posting cycle, which suggests the organization is scaling rather than contracting — a context that often shifts review sentiment within six to twelve months. Until dated, role-specific reviews accumulate on a clearly identified Heron Power profile, the public record offers a low-resolution snapshot: generally favorable in aggregate, noisy in detail, and insufficient to separate signal from entity confusion.
Who Thrives Here and Who Burns Out
Public employee-review data for Heron Power is effectively absent from the research corpus. No Glassdoor, Blind, Levels.fyi, or Comparably entries were surfaced for the Heron Power entity specifically, and no founder interviews or internal communications describing cultural norms were provided. That silence is itself a signal: either the company is too small to generate a review footprint, or current and former employees have not posted publicly in meaningful numbers.
What the first-party board data does show is a hiring slate concentrated in senior hardware and systems roles: Staff/Senior Staff Firmware Integration Engineer ($170k–$240k), Senior/Staff Power Electronics Engineer ($120k–$200k), Electronics Design Engineer ($150k–$200k), Automation & Controls Engineering Manager ($170k–$220k), Zero G Talent's board found, plus a Head of Sales ($200k–$240k), Zero G Talent's board's data shows, and a Staff Product Manager, Business Applications ($180k–$200k), Zero G Talent's board's figures put. The median posted band across 17 salaried roles sits at $195k, according to Zero G Talent's board.
In the absence of employee testimony, the roles themselves outline the likely success factors. Candidates who have shipped regulated or safety-critical hardware (think EV charging, grid-tied inverters, or industrial motor drives) will recognize the integration burden implied by "Firmware Integration" and "Automation & Controls" at staff level. The power-electronics band tops out at $200k, below the firmware and automation ceilings, which may indicate the organization values system-level integration over component-level optimization. The product-manager role, scoped to business applications rather than core hardware, hints that internal tooling and data pipelines are a lever the company is actively investing in.
Burnout risk, by inference, clusters around the same integration surface. Engineers who prefer well-scoped module ownership will struggle where firmware, power stages, and control loops must be co-optimized in the same sprint. The manager title on the controls role implies a small team (likely two to four reports), so the span is narrow but the technical depth expected is high; a lead who stops writing code or reviewing schematics will become a bottleneck fast. The sales role carries a $40k variable band ($200k–$240k), signaling quota pressure on a product that likely has long pilot cycles and utility-procurement timelines.
No review data means no verified attrition numbers, no work-hour patterns, no promotion-velocity signals. Candidates should treat the silence as a diligence item: ask hiring managers for the last three hires in each function and what surprised them at 90 days; request a conversation with a current IC on the firmware and power teams; probe how design reviews are structured and how often hardware spins. The factory rises in 2027. Until then, the job spec is the only contract on offer.
Working in frontier tech? Zero G Talent tracks the openings: see every open Heron Power role, browse frontier tech jobs, the companies hiring, and the people building the field.