Clearance Can Beat Code in Foundation Industries’ Hiring Funnel

The Roles Foundation Industries Is Filling

A robotics startup founded in 2024 has already deployed humanoid machines in combat, secured $24 million in U.S. military research contracts, CNBC reported, and set an 18-month timeline to field units with American forces. Foundation Industries needs to hire fast (its roadmap just leapt from prototype to pilot in Ukraine).

CEO Sankaet Pathak leads the company alongside co-founders Arjun Sethi and Mike LeBlanc. Their Phantom series targets a specific military problem: logistics in hazardous zones where human exposure carries unacceptable risk. MK-1 units, tested in Ukraine with U.S. government backing, carried roughly 44-pound payloads but lacked waterproofing and battery life for sustained deployment. Pathak says the Phantom 2 will double that payload and arrive with what he calls "superhuman abilities." Government conversations have moved from feasibility research to scaling discussions across Army, Navy, and Air Force channels.

That shift — a combat-tested prototype becoming scaled military hardware in under two years — shapes every role the company fills. Research contracts cover inspection, logistics, and weapons handling. The Phantom 2 cycle demands mechanical engineers who can solve payload-to-weight ratios while adding environmental sealing. The autonomy stack needs software engineers who build decision-making systems that retain human confirmation for most weaponized functions but execute fully autonomous actions in time-critical scenarios. Manufacturing engineers face Pathak's directive: deliver "the best robots we can build" to the U.S. military faster and cheaper than Chinese competitors can field their own humanoid programs. Every one of these roles passes through a single rigorous gate: Foundation Industries' screening process, which filters for candidates who can deliver under hard deadlines.

Some question the humanoid form factor itself. Melanie Sisson of the Brookings Foreign Policy program argued that Ukraine has shown the opposite need (rapid, cheap adaptation over complex human-like engineering). Toby Walsh, chief scientist at the University of New South Wales's AI Institute, expects tracked, flying, and underwater robots to replace human forces instead. Every role Foundation fills serves a technical milestone that validates or falsifies the humanoid thesis in military contexts.

The company's political alignment — including Eric Trump as chief strategy advisor and scrutiny from Senator Elizabeth Warren over government contract awards — adds operational complexity. Security clearance eligibility, export control compliance, and defense acquisition fluency become de facto qualifications regardless of the formal job description. Foundation builds a defense contractor that happens to be a robotics company.

The hiring imperative runs toward hardware milestones with hard deadlines. The next section examines how the company filters candidates against those milestones, and where the screen catches engineers who can actually deliver.

How Candidates Get Screened Out

Most applications die here. After submission, a first-pass review — human recruiter, applicant tracking system, or both — acts as a binary gate. Does the candidate meet minimum qualifications for the role? Yes or no. No, and the candidate is gone. Yes, and the next question follows: which preferred qualifications does this candidate hold?

This two-tier filter (minimum first, preferred second) structures the initial screen at most frontier-tech employers. Minimum qualifications are non-negotiable: citizenship or clearance eligibility for defense-adjacent work, specific degrees for engineering roles, years of experience with named toolchains or languages. Preferred qualifications differentiate: flight-software heritage, particular satellite bus experience, published work in relevant journals, prior exposure to regulated environments like ITAR or DO-178C.

Automation now does much of the work. Systems ingest, parse, and score resumes against the job description, ranking candidates by qualification relevance and producing a shortlist. Generative AI helps recruiters draft the descriptions that define those metrics. AI agents — software that understands goals, makes decisions, and acts autonomously across multiple steps and tools — are beginning to orchestrate the workflow from requisition to shortlist. Chatbots handle candidate FAQs and scheduling; some platforms deploy conversational agents that run the entire pre-screen for high-volume roles via text.

The shift from assistive AI (suggestions, recommendations, summaries) to agentic AI adds new power. Agents interact directly with candidates, perform multi-step tasks without human control at every step, and integrate deeply with CRM and ATS platforms to orchestrate end-to-end workflows. Beamery breaks roles into tasks; Eightfold draws on a proprietary global data set of more than a billion talent profiles to generate build-buy-borrow recommendations. Phenom's X+ Agents execute detailed workflows across the talent lifecycle under human supervision, and its Unified Orchestration Engine combines decision engines with simulations and human-in-the-loop governance.

These tools speed hiring and cut costs, but they introduce structural risks. Foundation models train on millions of job descriptions and resumes, along with outdated HR practices, biased language, and inaccurate information. The models don't produce identical outputs for identical inputs, and they hallucinate plausible-sounding but factually wrong content. Retrieval-augmented generation layers internal data onto foundation models, inheriting any quality issues in that data and potentially surfacing information buried in vector databases. When agents access multiple tools and data sources, organizations lose visibility into what data flows where and what decisions agents make autonomously, a phenomenon called autonomy creep.

Algorithmic bias remains documented. Datasets skewed toward accessible, mainstream groups create imbalances across gender and race. Existing social biases in raw data produce "bias in, bias out." When assessments consistently over- or underestimate a group's scores, they produce predictive bias. These discriminatory results are often overlooked because people assume AI processes are objective and neutral. Modern algorithms may appear neutral but can disproportionately harm protected class members (agentic discrimination by another name).

Responsible vendors are responding: iCIMS applies formal responsible AI principles around transparency, bias mitigation, auditability, and human-in-the-loop decisions. Employ's AI suite uses IBM watsonx.governance for transparency and bias monitoring. Phenom keeps human supervisors in the loop. But the regulatory and ethical burden lands on the employer deploying the system.

For a company in space and defense — where clearance requirements, export controls, and safety-critical engineering standards narrow the qualified population before the first resume arrives — the initial screen carries unusual weight. A false negative discards scarce talent; a false positive wastes expensive downstream interview cycles. The screen must be rigorous, auditable, and calibrated to actual role requirements, not the aspirational wish list that generative AI might produce from a vague prompt. Candidates who understand this filter, who map their materials to minimum and preferred qualifications, who anticipate the keywords and structures the parser expects, who treat the screen as a technical challenge rather than a formality, are the ones who advance.

What Clears the Bar — and Who's Hiring

Foundation Industries' screening funnel mirrors pressure across frontier-tech hiring: too many applicants for roles demanding a narrow intersection of technical depth, security eligibility, and program-scale experience. The company doesn't publish a rubric, but peer hiring data makes the priority stack clear.

SpaceX lists four principal-level roles with bands reaching $355,000, Zero G Talent found. Blue Origin adds two, banded up to $465,000. Thales Alenia Space recruits one in Irvine at $151,000–$252,000.

Ownership experience outweighs credentials. A candidate who led a propulsion test campaign at a launch provider, or shepherded a satellite bus from CDR to on-orbit checkout, carries more weight than a publication list. Board data shows companies paying a premium for "hands-on" qualifiers (SpaceX's Senior Software Engineer listing explicitly calls out "Hands on, Tech Lead") because the work demands debugging hardware in the loop, not simulating it in isolation.

Security clearance eligibility acts as a hard gate, not a preference. SpaceX's "Special Programs" roles in Washington, D.C., and Palo Alto sit inside classified envelopes; Blue Origin's TeraWave ground infrastructure and optical communication roles support a network architecture tied to national-security space. Foundation Industries operates in the same ecosystem. Candidates holding active TS/SCI clearance, or who can cross over without a sponsorship delay, move to the front. The screen often drops candidates who can't clear those requirements before a technical review begins.

Systems fluency across the hardware-software boundary is the third discriminator. Principal roles aren't siloed: Principal AI Engineer for Special Programs implies model deployment on radiation-tolerant compute; Principal NDE for Network & SDN implies test strategy for optical links surviving vibration and thermal cycling. Engineers who speak both flight software and RF link budgets, and have written requirements that survive both, clear the initial review.

Program-scale experience matters most at the director level. Blue Origin's Director of TeraWave Ground Infrastructure and Senior Director of Optical Communication carry bands above $440,000 because they require managing integrated schedules across contractors, government customers, and internal supply chains. Foundation's screening looks for the same scar tissue: candidates who have negotiated ICDs with a prime, defended a margin allocation in a Delta CDR, shipped flight hardware on a government timeline.

None of these qualifiers appear as a checklist in a job description. They appear in the resume of the person who gets the phone call. Publications, coursework, certifications matter less than the track record that lands the interview.

The sector's hiring surge amplifies the competition. Government investments in space exploration have grown steadily, from $67 billion during 2013–2017 to $81 billion from 2018–2022, reaching $93 billion for 2023–2027. That money flows to contractors and startups alike, inflating job volumes. In the past week, the sector's biggest players added hundreds of roles:

Ninety percent of U.S. employers now use AI screening tools to sort and rank job seekers, according to research from Stanford's Human-Centered Artificial Intelligence Institute. Most rely on the same few third-party vendors, creating a homogenized front-end hiring process that prioritizes keyword matching over human judgment. A study tracking more than three million job applications across 1,700 employers and 11 industry sectors found "substantial evidence of racial disparities in AI-based candidate screening." Black applicants faced discrimination at rates that, if corrected, would have advanced 40,000 more applications to the next stage. Asian applicants were affected at lower but still significant rates.

Foundation Industries' process combines technical assessments with structured interviews rather than the third-party AI screening vendors that dominate the broader market. This creates a tension: while most employers outsource initial filtering to tools that may introduce bias, Foundation Industries maintains more direct control over who advances. Research shows that even human-led processes can replicate systemic inequities when criteria aren't regularly audited, so the fairness question remains open.

The political dimension of AI hiring tools adds another layer. As chatbots and automated systems become more prevalent, concerns about ideological bias grow. A Washington Post analysis found that ChatGPT's underlying model answered nearly every political question with left-leaning arguments, while Google's Gemini took a both-sides approach in over 90% of responses. Companies like Anthropic and OpenAI have publicly committed to neutrality, though enforcement remains inconsistent.

Applicants in a Squeeze

The numbers describe a brutal funnel. Applications per job opening have doubled since spring 2022, and only 0.5% of applicants are ultimately hired, roughly 200 people compete for every position that gets filled. For job seekers targeting companies like Foundation Industries, this isn't a competitive market; it's a funnel that rejects almost everyone at the top.

The psychological toll shows up early. Recruiting expert J.T. O'Donnell told CNBC that applying online "has to be one of the most degrading and depressing things people do," and advised job seekers to stop applying through traditional channels. Job boards produce 61% of applications but only 42% of hires, so the math favors being recruited over applying. Referrals make up just 7% of applications yet account for 30–50% of all hires, and referred candidates are four times more likely to get hired than other applicants.

The system itself is buckling. Technical roles average 35–36 interviews and 26 interviewer hours per hire, and interviews per hire are up a third overall. Recruiters handle 93% more applications and manage 40% more open roles than in 2021, yet recruiting teams are 14% smaller. Hires per recruiter have dropped 43% since 2021. Job seekers feel it in slower response times, ghosting, and generic rejections.

Nearly two-thirds of job seekers report being ghosted after an interview, up 9 percentage points from 2024. Only 26% report having a great candidate experience. When companies layer AI screening on top of human overload, friction multiplies. Two-thirds of hiring managers use AI-detection software to screen resumes, while nearly four in five job seekers now use AI tools to generate their own applications. The result is an arms race: applicants optimize for algorithms they can't see, and employers build filters to catch AI-generated content.

The fairness question cuts deeper in high-tech hiring, where candidates are more inclined to attend AI-enabled interviews, yet job seekers generally view those interviews as less fair, because algorithms fail to consider interviewees' personalized expressions. AI systems "see" more but "signal" less, making applicants harder to influence through traditional impression management. Candidates feel scrutinized at a granular level but lack the knowledge to respond.

Still, job seekers adapt. Seven in ten employers now use skills-based hiring practices, and skills-based hiring can expand talent pools by nearly 16 times in the U.S. Nine in ten employers view non-degree certifications as important indicators of job readiness. But verification remains a bottleneck, as more than half of employers cite verifying skill claims as their main obstacle, and only 46% plan to expand skills-based hiring in 2026.

The counter-strategy is direct visibility. Instead of waiting for job boards to filter them out, successful candidates build presence: checking LinkedIn for posts from target companies, commenting thoughtfully, creating content, tagging companies, and connecting with employees. "You are literally creating a space where recruiters can find you and contact you," O'Donnell said, "and that's how you start getting interviews in this market."

Referrals remain the fastest path through the funnel. It takes 29 days to hire a referral versus 39 days for other sources. Referred hires show a 46% retention rate versus 33% for job board hires. Employers save roughly $3,000 per referral hire, and referral workers are 25% more profitable than non-referred workers.

Who Cares About This Screening

Foundation Industries' active recruitment sends ripples across at least three constituencies, the candidates competing for those positions, the company itself as it builds operational capacity, and the broader frontier-tech sector that watches hiring patterns as a signal of where capital and attention are flowing.

Applicants feel the most immediate impact. For applicants eyeing Foundation Industries' openings, standing out requires far more than technical skill; it calls for strategic vision, the ability to influence decisions, and a strong understanding of how businesses operate. Salary expectations add pressure. Space and defense companies post bands that give candidates a concrete yardstick: Thales Alenia Space lists roles at a median of $139,000 across 67 salaried positions; SpaceX posts a median of $150,000 across 1,279 roles; and Blue Origin's median sits at $183,000 across 1,046 positions. Foundation Industries' candidates will likely calibrate their expectations against these benchmarks.

Foundation Industries itself has a stake in how rigorous its screening proves to be. A process that is too permissive risks admitting candidates who cannot withstand the complexity of the roles. A process that is too restrictive risks losing talent to competitors who are also hiring aggressively. The company's screening choices also signal something about its culture and operational priorities to the outside world, which matters in a sector where the construction industry has been described as "highly lagged for digital transformation," marked by risk-averse norms, the requirement for specialized training, and the costs involved with applying high technology.

The frontier-tech sector at large watches these hiring patterns because they reflect where the industry is heading. When a company like Foundation Industries invests in a rigorous screening process, it reinforces a sector-wide expectation that defense-adjacent robotics demands people who can bridge the gap between advanced autonomy and the hard physical constraints of fielded machines, a challenge where the actual integration of these technologies into day-to-day project workflows determines whether a humanoid platform moves from a contract award to a convoy in the field.

Foundation Industries' screening process doesn't just sort candidates; it decides which engineers get to stand beside a Phantom 2 as it crosses from contract award to convoy in the field. For those who clear the filter, Pathak's "superhuman abilities" become their daily work. For the thousands who don't, the funnel closes with a parsed resume and a silent rejection.

Working in space? Zero G Talent tracks the openings: see every open Thales Alenia Space role, browse space jobs, openings at SpaceX and Blue Origin, and the people building the field.