Defense Tech Has $49 Billion. 9 Mothers Has 8 Roles. Engineers Are Scarce.

The Eight Open Roles at 9 Mothers

The counter-drone problem doesn't wait for hiring cycles. 9 Mothers has opened eight salaried roles at once, spanning the full stack from audio ML to the flight line, all onsite in Austin. The company's founders, Roman Khomenko, Russell Smith, and Bogdan Pyzh, graduated Y Combinator's P26 batch with a working system and paying customers, having already shipped $1.6 million of hardware to the Department of Defense with three more products queued behind EDDA.

The board data tells the current story: as of this week 9 Mothers shows eight salaried openings with a salary band of $150k–$400k (median $275k). Three roles were added in the past seven days. The newest additions signal where the bottlenecks are hardening:

Zero G Talent reported the Embedded Engineer (BSP), Staff band at $250K–$300K/yr. Zero G Talent's data shows the Perception Engineer, Staff band at $250K–$300K/yr. Zero G Talent found the Computer Vision Engineer, Senior band at $150K–$250K/yr. Zero G Talent's figures put the Machine Learning Engineer, Senior band at $150K–$250K/yr. According to Zero G Talent, the Robotics Engineer, Senior band is $150K–$250K/yr. Zero G Talent reported the Senior Mechatronics Engineer band at $150K–$250K/yr. Zero G Talent's data shows the Lead Mechatronics Engineer band at $250K–$400K/yr. Zero G Talent found the Robotics Engineer, Staff band at $250K–$400K/yr.

The company's public careers page lists six roles including Perception Engineer, Manufacturing Engineering Lead, Senior Optomechanical Engineer, UAS Pilot, Robotics Software Engineer, and Chief of Staff, while the board's live feed shows the eight salaried openings above. The company's stated philosophy — "short loops, real telemetry," "operators are users," "mass, setup time, and human factors are first-class specs", reflects a focus on rapid iteration. EDDA is the first product; the three follow-on products in various stages of development represent the next phase.

What Gets Past the Resume Screen

The average corporate posting draws 250 applications (Glassdoor). Roughly 75 percent never reach a human (Jobvite). At 9 Mothers, where eight salaried roles sit open in Austin, the funnel is tight. The company's board page shows that same band and median. That compensation attracts volume. Volume triggers automation.

Over 97 percent of Fortune 500 companies now run Applicant Tracking Systems before a recruiter opens a single file (ResumeAdapter). Harvard Business School research puts the ATS rejection rate at 75 percent. TheLadders measured the average initial scan at 7.4 seconds; other studies clock it at six (Ladders, 2018; SHRM). In that window the parser hunts for exact keywords from the job description. Missing one is the number-one cause of instant rejection (ResumeAdapter). Fancy templates (columns, tables, text boxes, graphics) break parsing. Image-based PDFs read as blank pages. Creative section headers like "My Journey" or "What I Know" confuse the algorithm (ResumeAdapter). The system scores what it can extract; garbled dates or misread job titles drag the score down. Knockout criteria are binary. Fail one, and the automated email fires before a person sees your name.

AI screening has moved from optional to standard. Aptitude Research found 65 percent of ATS vendors include it by default as of 2025 (Hyring). SHRM reports 44 percent of organizations use AI specifically to review or screen resumes (Testlify). The speed gap is stark: AI processes 1,000 resumes in the time a recruiter reviews ten (Hyring). Ashby's productivity data shows applications per hire rose 182 percent versus 2021 (Q4 2023–Q3 2024) (Testlify). Manual screening still burns 23 hours per hire (SHRM). The pressure to automate is real. So is the risk. Amazon's 2018 model penalized resumes containing "women's" (Hyring). The EEOC has warned that automated decision systems aren't automatically neutral (Testlify). Reuters has documented litigation alleging bias in AI-driven hiring tools (Testlify). NIST's AI Risk Management Framework is the reference for ongoing auditing (NIST).

When a human finally looks, they spend about six seconds on the first pass. Eye-tracking studies show 80 percent of attention lands on six areas: current role, current company, previous role, previous company, education, and skills (ResumeHeatmap). The decision is binary — reject or read more carefully. Recruiters compare candidates who cleared stages one through three. They hunt differentiators. Specificity wins: "Increased conversion 23 percent through A/B testing" beats "Improved metrics." Relevance wins: the candidate whose experience most closely mirrors the role. Recency wins: recent, relevant experience outweighs older accomplishments. Credibility signals win: recognizable companies, certifications, industry-specific proof (ResumeHeatmap).

For 9 Mothers, credibility signals map to defense tech. The open roles demand embedded systems, perception, computer vision, machine learning, robotics, mechatronics. A resume that lists relevant keywords (Python, C++, ROS, Linux, real-time systems, sensor fusion, CUDA, TensorRT, ONNX, flight software, DO-178C) in context with measurable results passes the keyword gate and the human scan. A resume that buries those terms in a paragraph of corporate-speak does not. The screen isn't magic. It's a filter for signal density.

The Technical Interview: Skills in Demand

9 Mothers' careers page states it is "hiring across audio ML, DSP, systems, and software" (9 Mothers). A Senior Applied Audio ML Engineer is expected to "develop machine learning models for acoustic target detection and classification," "bridge the gap between traditional DSP and modern ML," and "deploy models to edge runtimes with strict latency constraints" (Jaabz). The required stack is explicit: Python, Rust, TensorRT, ONNX, quantization (Jaabz).

DSP fundamentals are not optional. Published DSP interview questions cover the class of problems relevant to edge acoustic detection: optimizing FFT implementations for real-time systems, designing adaptive noise cancellation for non-stationary environments, debugging fixed-point overflow in filter chains, and comparing IIR versus FIR trade-offs when cycle count and phase linearity both matter (HelloIntern; ClimbTheLadder). One published question, "How would you design a DSP system for removing noise from a communication signal?", maps to the front-end of an acoustic pipeline (HelloIntern). Solutions cited in the literature (spectral subtraction, adaptive filtering, wavelet transform, ML-based enhancement) are fair game (ClimbTheLadder).

The company's own materials note: "We have multiple contracts, sales, etc." and "3 more products with demand in various stages of development" (ChReporter). The interview filters for engineers who can keep iteration loops short while the hardware is still moving.

Hardware on Software Tempo: The Cultural Fit

The phrase "hardware on software tempo" appears on 9 Mothers' careers page and in its Y Combinator hiring announcement — not as a tagline but as a constraint. The company builds an AI-powered counter-drone turret, EDDA, designed for close engagements, mobility, and practical deployment with qualities like low power, low cost, compact design, passive operation, and high-speed tracking (ChReporter).

The tempo mismatch is the origin story. Legacy counter-UAS systems were designed for fixed-wing intruders at altitude; acquisition cycles for those systems run a decade. FPV drones rewrote close-range warfare in two years. They are cheap, fast, expendable, and increasingly autonomous. The threat iterates in weeks. 9 Mothers exists because the procurement system cannot close that gap (9 Mothers). As the company puts it: "If you've ever shipped hardware inside a slow program of record and thought we should be doing this in months, not years — this is the team for you" (9 Mothers).

That sentence functions as a filter. It selects for engineers who have felt the friction of traditional defense programs, requirements freeze, change-control boards, years between flight tests — and want to operate on a different clock. The current team of nine (per BuiltIn) includes founders Russell Smith, Roman Khomenko, and Bogdan Pyzh. Smith's background at Rainforest QA, where he worked on product judgment and engineering rigor in a SaaS context, informs the cultural baseline: software-grade iteration applied to kinetic hardware (ChReporter).

End-to-end ownership is the cultural mechanism. "We own perception, tracking, and engagement end-to-end so the kill chain is one system — not three vendors and an integrator" (9 Mothers). The urgency framing also filters for mission alignment. The company describes its product as "built so operators can field it now — not after the next budget cycle" (9 Mothers). That language attracts people who have worked with end users — soldiers, not program managers, and know the cost of delay in lives, not milestones. It repels candidates who view defense tech as a funding vehicle for novel research. Observers note that more founders are moving toward categories where the stakes are higher, the engineering is harder, and the impact can be more immediate (ChReporter). 9 Mothers applies the same standard to its hires.

Practically, the culture means short feedback loops: hardware spins in weeks, not quarters; perception models retrain on field data from the last range day; the turret's software stack updates over-the-air. Candidates who need a stable requirements document before writing code will not last. Candidates who have built and flown their own hardware, crashed it, and rewritten the controller in a weekend — those are the ones who pass the cultural screen.

The Defense Tech Talent Market

Defense technology startups raised $49.1 billion in 2025, nearly doubling the $27.2 billion from 2024 and marking the sector's best funding year on record (Defense News). Deal counts rose 32.9 percent and funded-company counts increased 39.7 percent year over year (JobsWithDOD). Europe emerged as a growth-capital market, with startups raising $7.4 billion as of September 20 and forecasts projecting $10.5 billion by year-end (JobsWithDOD). Thirty-one defense tech startups now carry valuations of $1 billion or more (Startupill). Anduril Industries, the most valuable venture-backed defense startup, closed a $5 billion Series H in May 2026 that pushed its valuation to $61 billion and now employs more than 6,200 people after adding over 1,000 in nine months (Startupill; KORE1). Palantir sits above $250 billion as a public company (KORE1). Manufacturing-focused investment climbed to $4.7 billion across 39 deals in 2025 from $2.6 billion across 24 deals in 2024 (Defense News). The capital is converting directly into headcount.

The talent market has tightened accordingly. Roughly one in four senior backend candidates screened in 2026 is actively looking at a defense tech role, up from one in fifteen in 2022 (KORE1). Every senior backend, ML, and embedded engineer that Anduril hires is one fewer for everybody else, and the compensation those companies pay is recalibrating expectations across the entire industry (KORE1). At the frontier — Anduril, Palantir, Shield AI, senior engineers earn $300,000 to $500,000-plus in total compensation (KORE1). Cleared engineers command premiums of 40 to 100 percent over traditional defense contractor baselines (KORE1). A TS/SCI clearance takes six to 18 months to obtain; DCSA's 90th-percentile processing time for a Tier 5 investigation now sits around nine months (KORE1). The clearance bottleneck is real, and the pipeline is not scaling fast enough (KORE1).

Geography concentrates the fight. El Segundo and Costa Mesa, California, form the gravity well (KORE1). Austin, Texas, has quietly become the third pole, and 9 Mothers sits there, advertising eight salaried roles with bands from $150,000 to $400,000 and a board median of $275,000. The company's openings mirror the skill set the broader market is chasing. Rust is now standard for the autonomy stack at Anduril, Shield AI, Saronic, and most of the next tier (KORE1). C++ remains dominant for legacy and embedded work. Python drives ML and tooling. Go powers backend services (KORE1). The defense tech sector pulls roughly the same archetype of engineer that high-end fintech, infrastructure-software, and applied AI shops want (KORE1).

Hardware engineers are scarce. The simultaneous demands of robotics companies like Figure AI, GPU infrastructure providers like CoreWeave, and defense tech companies have created acute scarcity in mechanical engineering, embedded systems, and manufacturing roles (KORE1). Ninety-two percent of disclosed-pay roles at Defense Unicorns offer remote work options, a sharp contrast to prime contractors that need on-site presence for classified work (JobsWithDOD). Engineers see their work fielded in months rather than years (JobsWithDOD). Mission framing matters more than it used to; engineers who previously chose companies like Anthropic or climate tech for mission alignment are now considering defense tech as an equally compelling path (KORE1).

The traditional primes (Lockheed, Northrop, RTX, General Dynamics) hire at much larger absolute volumes, but their pace is set by program cycles measured in years rather than venture series rounds measured in quarters (KORE1). If you are hiring against the frontier, the move is not to match Anduril dollar-for-dollar. You will lose. The move is to be honest about where your offer is competitive and where it is not, and to recruit candidates whose mission fit, location, or career trajectory makes the trade-off real for them (KORE1). 9 Mothers is betting that Austin, counter-drone autonomy, and a hardware-on-software tempo will attract engineers who want to ship fast and see their code hit the field.

Advice for Aspiring Candidates

The eight roles 9 Mothers is filling in Austin — those same disciplines, read like a microcosm of where the entire defense industrial base is headed. Every title sits at the intersection of perception, control, and real-time compute. If your background lives in only one of those domains, you need a credible story for the other two before you apply.

Start with the resume. Defense-tech recruiters consistently advise: lead with a common job title, your city and state, email and phone, and (if you hold one) your clearance level and polygraph status (Oracle Blogs). 9 Mothers works on counter-drone systems for the modern battlefield; even if a specific posting doesn't list a clearance as required, the work will likely touch classified programs within your first year. Put the clearance at the top. Strip the objective statement. Replace it with a one-line summary that maps your hardest technical problem to the EDDA stack: fast, low-cost, autonomous-capable point defense for the FPV drone era (9 Mothers).

Your project portfolio should demonstrate hardware-on-software tempo. The founders have said explicitly they ship on that same timeline (9 Mothers). A GitHub repo that shows a perception pipeline running on edge hardware carries more weight than a paper citing simulation results. If you've closed the loop from sensor to actuator on a moving platform, document the latency budget, the failure modes you handled, and the test regimen you ran. Quantify the iteration cycle: how many hardware spins per quarter, how you validated regression without a full integration lab.

Prepare for the technical screen by rehearsing the problems that appear in C-sUAS work: multi-target tracking under clutter, sensor fusion when the RF environment is contested, model compression so a detection network fits in constrained GPU memory while maintaining throughput. The Perception Engineer, Staff and Computer Vision Engineer, Senior roles will probe your grasp of classical estimation (EKF, UKF, particle filters) alongside modern deep-learning trackers. The Embedded Engineer (BSP), Staff interview will test board bring-up, DMA-driven sensor ingest, and deterministic scheduling on RTOS or bare metal. The mechatronics roles expect you to speak fluently about thermal management in a sealed enclosure that survives vibration and shock, and to show you've designed for manufacturability at attrition-scale unit costs.

Cultural fit at 9 Mothers means comfort with ambiguity and a bias toward shipping. The Y Combinator Spring 2026 batch accelerated their visibility, but the pace hasn't slowed. Candidates who thrive here treat requirements as hypotheses, not contracts. In conversation, reference the company's own language: "engineered for attrition," "software-tempo hardware" (9 Mothers). Ask how they validate autonomy regressions under range constraints. Ask what the current bottleneck is between the perception stack and the effector. The quality of your questions signals whether you've already mentally joined the team.

Finally, treat the defense-tech talent market as what it is: a seller's market with 11,000+ open roles across the sector (LinkedIn) and funding that grew more than tenfold from 2020 to 2025 (CNBC). You have leverage, but only if you can prove you'll reduce risk on day one. Tailor every artifact (resume, portfolio, interview narratives) to the specific problems 9 Mothers is solving this quarter. The roles are live on the board today.

The Future of Defense Tech Hiring

The company's salary bands, $150,000 to $400,000, match what the market now pays for engineers who can move algorithms onto hardware at operational tempo. But the implications stretch far beyond one startup's headcount.

The Pentagon has spent decades trying to close an AI talent gap that the National Security Commission on Artificial Intelligence called "one of the greatest impediments to the U.S. being AI-ready by 2025" (GAO). As of late 2023, the Chief Digital and Artificial Intelligence Office still had not assigned responsibility or set a timeline to define and identify the AI workforce across the department (GAO). The Office of Personnel Management missed its statutory deadline to create an AI occupational series, opting instead for classification guidance that agencies must interpret themselves (GAO). Eleven of sixteen DOD organizations interviewed by GAO cited the lack of a definition as the reason they had not begun developing AI workforce strategies (GAO). The bureaucracy that manages nearly 3 million employees cannot yet count how many of them do AI work (GAO).

Meanwhile, the money is moving. Global military spending hit $2.70 trillion in 2024, the sharpest yearly increase since the Cold War ended (JobsWithDOD). NATO allies plan to raise defense spending targets from 2 percent to 5 percent of GDP by 2035 (JobsWithDOD). U.S.-based defense tech startups pulled in roughly $38 billion through the first half of 2025, a pace that could exceed the 2021 peak (JobsWithDOD). The Pentagon's FY2026 budget request asks for $179 billion in RDT&E, a 27 percent jump (JobsWithDOD). IDC forecasts U.S. aerospace and defense spending on AI and generative AI will reach $5.8 billion by 2029, 3.5 times 2025 levels (Deloitte). Defense satellite launches are projected to grow 160 percent in the decade to 2034 (JobsWithDOD).

That capital demands people who can deploy it. Deloitte projects the share of industry job postings requiring data analysis skills will rise from 9 percent in 2025 to nearly 14 percent by 2028; data science skills from 3 percent to 5 percent (Deloitte). By 2026, agentic AI is expected to progress from pilot projects to scaled deployments across decision-making, procurement, planning, logistics, maintenance, and administrative functions (Deloitte). The focus is shifting from hiring AI specialists to embedding AI fluency across the entire workforce (Deloitte).

Throughput — the ability to produce, replenish, and sustain at pace, stops being a factory metric and becomes a national security input (Deloitte).

The constraint is not just headcount. It is the sub-tier supply chain where multi-tier visibility is limited, working capital is tight, and qualification timelines run long (Deloitte). A prime contractor cannot demand higher rates from a tier-3 casting supplier that lacks capital for automation, the workforce to add shifts, or required cybersecurity certifications (Deloitte). Certification timelines slow supplier onboarding and narrow the near-term pool of eligible vendors (Deloitte). The ecosystem faces a widening gap between what can be designed and what can be produced, replenished, and sustained at rate (Deloitte).

Structural responses are emerging. Five Defense Tech focused colleges launch in 2026 (JobsWithDOD). The Defense Skills Passport concept aims to let professionals move smoothly between armed forces, industry, and related sectors (JobsWithDOD). The Army created the 49B AI/ML Officer area of concentration; the Coast Guard established a Robotics Mission Specialist rating (JobsWithDOD). The NDAA allocated $30 billion to cybersecurity (JobsWithDOD). SkillBridge and Army Career Skills Programs connect transitioning service members to training in their final 180 days (JobsWithDOD). Universities partner with contractors (Saab with Purdue, MIT, and Syracuse) to tailor curricula (JobsWithDOD). Employers sponsor clearances; candidates cannot apply independently (JobsWithDOD).

On the data side, modern AI-enabled platforms like Workday Government can ingest unstructured data (project histories, role descriptions, resumes, performance reviews) to build dynamic skills inventories (DefenseScoop). Commanders can instantly search for hidden talent: a logistics officer with advanced data science skills, ready for an urgent mission role (DefenseScoop). The platform has FedRAMP Moderate authority; Impact Level 4 for most of DOD and an Air Gap region for top-secret data are imminent (DefenseScoop). The alternative (retrofitting customized HR systems or layering federation software on legacy stacks) reproduces the technical debt that slows every other defense modernization (DefenseScoop).

The same profile — hardware-fluent, software-fast, clearance-ready, will be the scarcest resource in a defense industrial base that must scale throughput while the workforce ages (60 percent of employees are 35 to 54) and women remain 27 percent of the talent pool (MilitaryDispatches). The startups that crack the hiring loop now, and the primes that restructure their sub-tier relationships around it, will set the pace for the next decade. The rest will be buying capacity they cannot fill.

Working in frontier tech? Zero G Talent tracks the openings: see every open 9 Mothers role, browse frontier tech jobs, the companies hiring, and the people building the field.