Six-Person Hologram Labs Pays $185K Median for Just 3 Roles

Three Roles. One Screen.

Hologram Labs shows exactly three salaried openings on the Zero G Talent board: a Revenue Operations Manager, a Solutions Engineer II, and a Sales Engineer, all United States-based. One role appeared in the past week. Founders Hongzi Mao and Tong Pow lead a team still in the single digits.

That three-role slate is the entire hiring window. No pure research scientist. No infrastructure engineer. No product manager. The titles reveal the priority: two roles sit at the revenue-technical interface, the third builds the machinery that lets a tiny go-to-market motion scale without a sales army. Hologram buys distribution leverage and technical validation capacity, not raw model-building headcount.

The pattern mirrors what SignalFire tracked across twelve "Tech Majors" in 2025: engineers made up 55% of new hires, but the middle-management layer (the people who translated strategy into roadmaps) shrank 41% as AI absorbed that work. The org chart didn't shrink evenly; it flattened. Generating code got cheap. Judging it did not. For a lab at Hologram's stage, that flattening means every hire must carry disproportionate judgment load. That position isn't a junior support role; the "II" signals someone who can scope a deployment, debug a customer integration, and feed product signal back to the core team without a product manager in between. The Sales Engineer carries similar weight: technical credibility in front of buyers, not demo scripts. That role builds the systems that let two founders and a handful of engineers run a revenue motion that would have required a dozen hires five years ago.

External boards list other titles (Applied ML Engineer, 3D Art Director, Product Designer), but those postings either predate the current window or live on platforms where roles linger after they're filled. The Zero G Talent board ingests directly from the source; its three roles are the live surface of Hologram's actual hiring intent right now.

Wage-premium data explains the salary bands. PwC's 2026 Global AI Jobs Barometer puts the AI-skills premium at 62% over comparable non-AI roles, up from 56% a year earlier and 25% the year before. ManpowerGroup's 2026 survey of 39,000 employers across 41 countries ranked AI skills the hardest to hire for globally, ahead of engineering, IT, and skilled trades. Hologram's bands sit inside that market reality: they pay for the judgment layer AI hasn't commoditized.

Three roles. One screen. The next section breaks down what that screen filters for.

What the Screen Filters For

Hologram Labs publishes no hiring playbook. What exists instead is a pattern visible across its few public touchpoints: a founding team that built its reputation on reinforcement learning and distributed systems, a "Founder Intern" role designed to bypass the slide-deck internship cycle, and three go-to-market positions sitting unusually close to the product for a company this small. The screen extends the founders' technical filters: Hongzi Mao's work in blockchain-based AI and Jiachen Mao's product-and-engineering remit suggest a bias toward candidates who have operated inside the same constraints the lab faces now.

First-party board data shows the same three salaried roles open this cycle, all United States-based. That concentration — no pure research roles, no junior ICs, no generalist "founding engineer" catch-all — functions as a filter before a human reads a résumé. Candidates who apply without a clear map to one of those three lanes stall.

A "Founder Intern" post on LinkedIn offers the clearest signal of what the lab values. The role was framed as the antithesis of "12 weeks of slide decks, silent standups, and a return offer to do the same." That language (public, unprompted) doubles as a negative spec: no performative process, no hierarchy for hierarchy's sake. By implication, applications leading with brand-name internships or conference talks but no shipped artifact in a relevant domain (distributed systems, RL environments, on-chain inference) get filed.

Positive signals are narrower. The Solutions Engineer II and Sales Engineer roles demand technical fluency deep enough to sit with a customer's engineering team and debug integration issues in real time. In each case, the screen rewards evidence the candidate has already operated at that scope.

No formalized rubric exists: no published scorecard, no leaked interview loop, no on-the-record quote from Tong Pow or Hongzi Mao describing the exact gate. The "Innovation Hologram" platform described by Simeon Lee, which markets video intelligence and behavioral analysis for recruiting, is a separate venture unconnected to Hologram Labs' own process. Absent documentation, the screen is best read through the roles themselves: three slots, each with tightly bounded scope, each priced for someone who has already done the work. The lab buys proof that the candidate's last hard problem looks like Hologram Labs' next one.

Proof of Work, Not Pedigree

The frontier-lab playbook has settled on a single, verifiable artifact: a 100-word claim backed by a live artifact and references who can independently describe the same decisions and working style. Hologram Labs applies that standard, but narrower: candidates who clear the screen don't just describe a project; they show a shipped product that reached real users and, ideally, generated revenue.

The lab's own founder-intern program makes the expectation explicit: three builders spent a month in a Los Angeles hacker house, each taking an idea from concept to App Store release, running their own ad campaigns through AVA Studio, and racing to land the first dollar of revenue. Weekly progress reviews with the full team replaced mentor-office-hours theater; the interns built alongside the core staff, not adjacent to them.

Applicants who internalize that model submit links to live URLs, not GitHub repos with stale commits. References matter only when they corroborate the artifact. The strongest signals come from environments where the candidate had no safety net, no dedicated infra team, and no excuse for "it works on my machine."

The board shows three salaried roles open this month, each carrying a compensation band reflecting the expectation of immediate, autonomous contribution. The founding team has said publicly that AVA Studio exists to collapse the idea-to-revenue loop for builders; the hiring screen is the same loop applied to talent. Candidates who treat the application as a pitch deck get rejected. Candidates who treat it as a demo reel — live product, measured impact, named references who watched the decisions happen — move forward.

Why the Window Stays Narrow

Hologram Labs' hiring slate is tiny because the company itself is tiny — and deliberately so. As of October 2026, the lab lists six employees on its cap table, a figure aligning with the three salaried roles currently posted. The headcount ceiling is the real constraint. Every hire must earn its keep in a nine-person organization still proving its core technology at scale.

The funding timeline explains why the ceiling hasn't risen. The company raised a seed round on July 14, 2022, led by Polychain Capital with participation from Nascent, Inflection, The Operating Group, Quantstamp, Neon DAO, Foothill Ventures, and South Park Commons, plus angel checks from Mike Shinoda of Linkin Park, Kalos of Parallel NFT, Richard Ma of Quantstamp, and Naveen Jain of Yat Labs. That round valued the company at a post-money figure PitchBook marks "Generating Revenue," a signal the business model worked, but the check size was modest by AI-lab standards. A second seed tranche, labeled Seed VC-II, closed January 1, 2026, with MH Ventures; cbinsights tallies total capital raised across six rounds. In an era where frontier AI labs routinely announce nine-figure Series A rounds before product launch, Hologram's cumulative raise is a hard budget constraint. The founding duo — CEO Tong Pow, formerly of 0x Labs, and CTO Hongzi Mao, an MIT computer-science PhD — chose to extend runway rather than bloat headcount.

Product architecture reinforces the constraint. Hologram's flagship is a machine-learning model compressed to run in real time inside a browser, delivered as a Chrome extension that swaps a user's webcam feed for a motion-tracked NFT avatar. The model ingests face-tracking points from the camera and drives lip-sync and head movement on the fly, with no GPU cluster, no cloud inference, and no heavy client install. That technical choice means the engineering surface area is narrow but deep: browser-based ML optimization, WebGL/WebGPU pipeline work, and real-time computer vision that must hit 30+ fps on consumer laptops. The team simultaneously builds hand-tracking support, a mobile application, and tooling for short- and long-form video creation. Each thread demands specialists who can ship production code in a codebase that cannot afford abstraction leaks.

Revenue exists but concentrates in a niche. The company counts over ten NFT collections (Cool Cats, Deadfellaz, Crypto Covens among them) as paying partners for avatar animation services. That commercial traction validates the go-to-market motion, yet the revenue base remains tied to a crypto-native segment. The founders have explicitly said crypto is optional for the technology; the model works on any static avatar, blockchain-verified or not. But the current sales motion, reflected in the open Solutions Engineer and Sales Engineer roles, still orients toward NFT collection integrations and creator workflows. Expanding beyond that wedge requires product maturity (hand tracking, mobile, video export) the current headcount is racing to deliver.

The result: a hiring window admitting only candidates who can contribute immediately to one of those three threads. A generalist ML researcher, a backend engineer without browser runtime experience, or a sales hire without technical credibility will not clear the screen, not because the bar is arbitrarily high, but because the organization cannot absorb onboarding time. The narrow window is a direct function of a six-to-nine-person team, a war chest, and a product that only works if the model stays small enough to run in a browser tab.

What This Means for Frontier AI Job Seekers

The hiring screen at Hologram Labs is not an outlier. It is a leading indicator of how small, capital-efficient AI labs filter a flood of AI-polished applications. Greenhouse data shows the average opening attracted 244 applications in 2025 (more than double the 2022 volume), while time-to-offer stretched to 47 days as of November 2025. Hiring managers now spend roughly half their week filtering AI-generated spam. The old signals — clean resumes, polished cover letters, keyword-matched skills — have collapsed.

Candidates targeting labs like Hologram Labs need to invert their approach.

First, stop spraying generic applications. Labs running narrow slates (three roles, one screen) don't scan for keyword coverage. They hunt for proof of work mapping to their immediate technical constraints. Read the job description for the actual problem, not the wish list. Scion Technical's analysis of AI hiring strategies emphasizes defining business outcomes first, then mapping to technical capabilities. A candidate who fine-tunes foundation models may be the wrong hire for a lab needing productionized pipelines in a regulated environment. The distinction matters more than the stack.

Second, prepare for practical, timed assessments, often screen-recorded. The best assessments aren't the longest; they're the most relevant. They test deployment judgment, trade-off thinking, and the ability to partner across product, engineering, security, and executive stakeholders.

Third, demonstrate learning velocity over framework tenure. As framework releases accelerate, the ability to learn a new tool in a week outweighs five years on an older one. PeopleInAI's hiring roadmap puts it bluntly: "An engineer who adapts to new models is worth more than one who only knows one legacy system." Show it. Point to GitHub repos where you built tooling around a new architecture. Cite ArXiv preprints you implemented before they were packaged. Theoretical knowledge isn't enough; labs need framework proficiency (PyTorch, TensorFlow, JAX) paired with production evidence: how you launched models for users, handled latency, cost, governance, and downstream impact.

Fourth, be honest about AI use in your own application. Hiring teams increasingly ask: "How did you use AI in this process?" They don't penalize usage; they test for transparency and judgment. A candidate who discloses thoughtful AI assistance (drafting, then refining with domain knowledge) signals the human-AI collaboration skill that HBS research identifies as the key driver in augmentation-prone roles. One who claims pure manual effort while submitting AI-perfect prose raises a red flag.

Fifth, evaluate the lab as rigorously as it evaluates you. Retention risk begins during recruitment. Scion Technical warns that oversold roles, vague leadership expectations, and data environments less mature than presented drive attrition. Top candidates ask: Will I have executive support? Usable data? Production pathways? Authority to influence outcomes? If the answers are thin, the role is a trap, regardless of compensation. BCG data shows 80% of AI talent leave for more interesting work or stalled advancement; only 10% of new roles are filled internally. The tech culture expects promotion every 12–18 months. A lab that cannot articulate a clear advancement path will bleed talent.

Finally, target your search. The strongest experts hide in research communities and open-source projects, not job boards. Labs like Hologram Labs (founded by the duo, running a tight slate of three roles) don't need volume. They need precision. Candidates who treat each application as a tailored proof-of-work submission, aligned to the lab's stage and constraints, clear the screen. Those who don't, won't.

The browser tab stays open. The model stays small. The next hire ships in the same constraint.

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