Who Gets Hired
The emergency call comes in. A telecommunicator answers. Behind the scenes, software decides what gets surfaced, what gets routed, what gets missed. Aurelian builds that software. The company hires people who can make it reliable at the scale of a 911 center on its worst day.
This guide compiles the hiring bar, compensation ranges, interview loop, work environment, and who actually lasts at Aurelian, so candidates can prepare without guesswork.
As of October 2026, Aurelian lists 13 live openings: eleven in Seattle, one remote. The seniority split is even: six associate-to-mid roles, seven senior-to-lead. Not a single posting accepts candidates with fewer than three years' experience. The bar starts at mid-level.
Engineering dominates. RoleIQ tracks eight engineering openings; Zero G Talent shows three staff-level engineering postings (Staff Product Engineer, Staff Infrastructure Engineer, Staff Backend Engineer) all in Seattle, plus a Senior Product Engineer and a Tech Lead/Engineering Manager role. A Founding Product Manager role and a Strategy role round out product and business functions. A single Recruiter opening signals the team is still building its own hiring muscle.
The skill clusters tell a clearer story than any job title. LESS and Qualification appear in twelve and eleven role descriptions respectively. Analytics, Usability, and Multimodal each show up in five or more. Cloud, Logging, and AI Agents trail but appear consistently. This stack is built for real-time audio processing, transcription accuracy, and low-latency decision support, not a generic SaaS stack.
Median days open: 52. Roles stay live long enough to filter hard. The Recruiter posting appeared one day before the October snapshot; the Founding Product Manager, six days prior. Engineering and sales roles have been open two to three months. A net loss of four openings over 28 days suggests either fills or pullbacks, but one new People/HR role appeared in that window, and RoleIQ recorded a fresh addition in the last week.
Candidates who clear the experience floor still need fluency with the domain constraints: public-safety regulations, audio pipeline reliability, the particular cadence of a PSAP floor. The careers page frames the mission as "making every call better for telecommunicators and the communities they serve."
What Aurelian Pays
Aurelian's compensation data comes from nine salaried roles posted to the Zero G Talent board, all based in Seattle. The aggregate band runs roughly $148k–$242k base, with a median around $240k, Zero G Talent's data shows. That median sits high because the sample skews toward staff-level engineering and management titles, roles that typically command the top of any startup's pay scale. Read the ranges below as the floor and ceiling advertised for each specific slot, not a universal ladder.
| Role | Base Salary Range (USD/year) | Location |
|---|---|---|
| Strategy | $150,000 – $250,000 | Seattle, WA |
| Tech Lead / Engineering Manager | $160,000 – $240,000 | Seattle, WA |
| Staff Product Engineer | $180,000 – $240,000 | Seattle, WA |
| Staff Infrastructure Engineer | $180,000 – $240,000 | Seattle, WA |
| Staff Backend Engineer | $180,000 – $240,000 | Seattle, WA |
| Senior Product Engineer | $150,000 – $200,000 | Seattle, WA |
The three staff engineering tracks — Product, Infrastructure, and Backend — share an identical $180k–$240k band. That parity signals the company values those disciplines equally at the staff level. The Tech Lead/Engineering Manager band overlaps the staff range at the top but starts $20k lower, reflecting the dual-track reality that a first-time manager may not yet carry the same scope as a seasoned staff IC. Strategy sits widest at $150k–$250k, as Zero G Talent's board shows, a spread that likely captures both senior individual contributors and early leadership hires for a function still being defined.
Senior Product Engineer is the only role with a ceiling below $240k, topping out at $200k, Zero G Talent's board found. The $40k gap to the staff band is the clearest market signal on the board: Aurelian treats the staff promotion as a meaningful step change in scope and compensation, not title inflation.
Equity details do not appear in any board posting or on the public careers pages. No option-pool percentage, strike-price methodology, or refresh-grant cadence has surfaced. Candidates should ask directly about the equity component (grant size, vesting schedule, whether the company uses standard four-year monthly vesting with a one-year cliff) before treating the base figures above as total compensation.
The Seattle-only tag on every listing matters. Aurelian does not publish geographic differentials, so remote or hybrid candidates outside Puget Sound should clarify whether bands adjust for cost of labor or stay flat. The company describes building "purpose-built technology for PSAPs" (public safety answering points), which suggests a product that may require on-site collaboration with emergency dispatch centers. That operational constraint could explain the geographic concentration.
Nine data points is a thin sample. The board shows no entry-level, mid-level, or non-technical roles, so the ranges above represent the upper half of the org chart only. As the company grows, expect the band to widen downward. The equity piece remains the unknown variable.
How the Interview Loop Works
The interview loop is short by design: a recruiter screen followed by one or two technical rounds, but the bar inside those rounds is high. Interviewsense data confirms the sequence: an initial conversation with a recruiter, then 1–2 technical sessions blending data-structures-and-algorithms problems with system-design discussion and behavioral probing. The company's own guide describes the philosophy as "problem-solving in context": no gotcha trivia, no LeetCode regurgitation. Instead, interviewers present scenarios mirroring the actual work: extracting structured data from a conversational interface, persisting it as auditable business records, exposing deterministic REST endpoints alongside LLM-driven chat tools.
Recruiters screen for three things up front: fluency in at least one modern backend language (Go, Java, Python, or C++), provable experience building or maintaining distributed systems, and a deep grasp of data structures, algorithms, and complexity analysis. Cloud-native infrastructure (AWS, GCP, Kubernetes), high-scale environment exposure, and observability/SRE tooling are listed as nice-to-haves. The stack you'll actually touch — FastAPI, SQLAlchemy async ORM, SQLite via aiosqlite, Alembic migrations, OpenAI Python SDK, Pydantic v2, pytest/httpx on the backend; Next.js 14 App Router, React 18, TypeScript, Tailwind, SWR on the frontend — matters less than the architectural reasoning behind it. Candidates who can explain why the backend owns system-of-record data, why tool calling bridges free-form intent to typed actions, and why server-side validation remains mandatory even when the model "already collected the data" signal they've internalized the mental model the team works from daily.
A strong application demonstrates that reasoning before the first call. The take-home assignment, extended from a narrow assessment into a full CRUD workflow with revision history, is the clearest window into what Aurelian values. The original spec asked for three pieces: a FormSubmission row created when the model calls submit_interest_form, REST endpoints to update/list/filter/delete forms with status validation (None, 1, 2, 3), and a generic change-history design tracking field-level before/after values and timestamps. The current repository contains all three implementations plus a frontend history modal, chat-driven deletion, and extra validation. Interviewers will ask you to walk through your choices: why JSON diff instead of full snapshots (answer: the question is "what changed," not "what was the whole state"), why two OpenAI calls in PUT /chat/{chat_id} (first decides tool use; second produces the natural-language response informed by tool results), why a generic history table (the requirement explicitly anticipated future entities needing revision tracking), and why both REST CRUD and chat-tool CRUD coexist (deterministic UI actions vs. conversational workflows). Candidates who treat the take-home as a checklist and cannot articulate those trade-offs rarely advance.
Preparation data from the company's guide suggests most successful candidates spend two to four weeks focused on the loop. The guide explicitly warns against memorizing solutions ("recycled answers are often easy to spot") and urges candidates to verbalize assumptions, explain the "why" behind every decision, and treat the collaborative coding environment as their own IDE. The most common rejection reason, per the same source, is failure to articulate the reasoning behind technical choices. Interviewers are instructed to provide hints and steer the conversation if a candidate stalls on a syntax detail; they are evaluating thought process, not perfect recall.
Behavioral and leadership assessment runs through every round, not just a dedicated slot. Even at the individual-contributor level, the rubric looks for signs of technical leadership: driving consensus, translating complex ideas for non-technical stakeholders, advocating for engineering best practices. Values alignment centers on ownership and curiosity: how you respond to feedback, whether you demonstrate a growth mindset when facing unfamiliar problems. The guide's "Final Mental Model" summary — chat as conversation container, messages as raw transcript, FormSubmission as structured record, PUT /chat/{chat_id} as orchestrator bridging LLM behavior to database writes, REST for deterministic UI, ChangeHistory for auditability, Alembic for reproducible schema evolution — is essentially the cheat sheet interviewers expect you to reconstruct from first principles. If you can't, the loop ends early.
Where the Work Happens
Every open role on Zero G Talent's board lists one location: Seattle, Washington. That concentration tells you how the company operates: the team builds, deploys, and supports its 911-call automation platform from a single hub rather than a distributed network of offices.
Aurelian's product — an AI voice assistant that handles roughly three in four non-emergency 911 calls without dispatcher involvement — sits at the intersection of real-time audio processing, telephony integration, and compliance-heavy public-safety procurement. The board's salary bands reflect that blend of systems and ML expertise.
The company blog describes its technology as "purpose-built for PSAPs" and emphasizes that the system "gathers essential information up front, triages call urgency, and provides status updates to callers." Aurelian's $14 million Series A, as the company blog reported, funds the team that runs those drills. The funding post notes the AI "gives trained professionals the time and bandwidth to focus on high-priority emergencies and complex coordination." That mission, reducing dispatcher burnout while cutting wait times, requires close feedback loops with actual 911 centers. The blog references "helping Chiefs understand the opportunity" and working with agency leadership directly.
The research doesn't show satellite offices, a dedicated hardware lab, or a separate research campus. If those exist, they aren't reflected in current hiring data. What the board and the company's own writing confirm is a Seattle-centered team building voice-AI infrastructure for the hardest constraints in public safety: zero-downtime audio, regulatory audit trails, and human-in-the-loop oversight. The work happens where the telephony meets the model, and right now, that's Seattle.
Who Thrives Here
The people who last at Aurelian share a specific tolerance: they build software that answers the phone when someone might be dying. Since launching in May 2024, the company's AI voice agent has gone live in more than a dozen 911 centers (Snohomish County, Washington; Chattanooga, Tennessee; Kalamazoo, Michigan), fielding thousands of real calls daily. That production footprint is the filter. NEA partner Mustafa Neemuchwala, who led the $14 million Series A announced in August 2025, said: "As far as we know, nobody else is actually live." The bar isn't theoretical. A misrouted non-emergency call is a nuisance; a misrouted emergency call is a headline. Engineers who ship here have already internalized that distinction.
The technical stack reflects the constraint. Voice AI over legacy telephony, real-time transcription, and emergency-classification models that must transfer to a human dispatcher in milliseconds are not sandbox problems. The open roles on Zero G Talent's board tell the story: staff and senior engineering tracks clustered at $180k–$240k. That compensation buys engineers who have operated production systems where downtime carries severe consequences.
If your background is consumer SaaS feature factories, the on-call rotation will feel different. The company describes the work as triaging noise complaints, parking violations, stolen wallet reports so human dispatchers can "have a chance of taking a break or go to the bathroom," in founder Max Keenan's words, framing the mission as capacity relief for a workforce running 12- to 16-hour shifts in an industry with top-10 turnover rates. You are building leverage for people who don't have enough of it.
That mission filter selects for a working style. The pivot from hair-salon booking automation (Keenan's original Y Combinator S22 idea) to 911 triage required discarding product code and retraining models for a domain where "good enough" doesn't exist. The team that made that turn did it without a public relaunch narrative; they shipped to dispatch centers and iterated on live traffic. Candidates who need polished specs before writing code will stall. The ones who thrive treat the first deployment as the real spec review. They also work through government procurement cycles: pilot agreements, security reviews, interoperability with CAD systems that predate the web. That process rewards patience that isn't passive: it's the patience of someone who has already built the integration in their head while the contract clears.
The culture signals are sparse in public record, but the hiring pattern is legible. Every open role is senior or staff level. There are no junior listings, no "new grad" bands. Aurelian hires people who have already led the hard part of a system (the part that fails at 3 a.m.) and want to do it again with higher stakes. The Seattle colocation matters too: the team sits near the cloud infrastructure they depend on. If you've shipped at major cloud providers and want your next pager duty to protect a dispatcher's lunch break, the signal matches. If you're optimizing for feature velocity and launch metrics, the mismatch will show in the first technical interview.
The emergency call comes in. The software Aurelian builds decides what gets surfaced. The people who get hired here are the ones who've already made that decision, in systems where the cost of error isn't a bug ticket — it's a headline. That's the bar. The rest is preparation.
Working in frontier tech? Zero G Talent tracks the openings: see every open Aurelian role, browse frontier tech jobs, the companies hiring, and the people building the field.