PlanetScale’s Top Offers Reach Near‑$1M Total Comp for Vitess Roles

The Surge in Context

PlanetScale posted 13 open roles this month, four in the past week alone on Zero G Talent's board. The breakdown skews heavily technical: seven software-engineering slots spanning Vitess internals, Postgres development, sharding orchestration, insights tooling, and internal platforms, plus three customer-facing solutions roles, one design position, and a sales development representative.

The headcount context sharpens the picture. PlanetScale sits at roughly 125 employees after raising $185 million, Fast AI Jobs reported, across four rounds, the last a $50 million Series C, Fast AI Jobs' data shows, in November 2021. That capital runway, combined with a product built on Vitess (the YouTube-born, CNCF-graduated sharding layer that already powers GitHub, Slack, and Etsy), means the company isn't hiring for hypothetical scale. The managed MySQL and Postgres platform, with its branching workflows and non-blocking schema changes, is already absorbing production workloads at companies that can't afford downtime.

What connects the funding, the Vitess lineage, and the 13 roles is a single variable: the database layer is becoming the AI infrastructure layer. PlanetScale's own careers page frames it directly: "the future of AI infrastructure: optimize and shard your database with agents." That positioning turns every engineering hire into a bet on whether the team can extend Vitess-derived architecture into the workload patterns that LLM-driven applications demand. The hiring boom isn't about headcount. It's about whether the systems thinking that scaled YouTube's MySQL can now scale the data layer underneath the next platform shift.

The Bar: Systems Fluency Over Keywords

PlanetScale's hiring bar doesn't bend for its remote-first posture — it sharpens. The company screens for a specific blend of systems fluency, ownership proof, and communication discipline that reflects what it actually builds: a managed database platform running on Vitess and a new Postgres sharding architecture called Neki. Recruiters and automated filters look first for job-description keyword alignment, credible ownership bullets, and eligibility: location, level, work authorization. A résumé that buries impact under tool lists rarely clears the initial scan.

That philosophy translates into concrete evaluation criteria. Early technical filters reward timed accuracy and clear communication of approach, not algorithmic party tricks. Interviewers probe fundamentals, system or domain depth, and how a candidate recovers from ambiguity — outline the approach before coding; if you stall, talk through trade-offs rather than go quiet. The hiring-manager and bar rounds confirm seniority and team match: candidates bring numbers that show what they changed, a conflict-resolved or project-salvaged story, and a specific "Why PlanetScale?" tied to the business unit (Neki Orchestration, Insights, or core Vitess) — not the brand.

The competencies map directly to the product. PlanetScale powers production databases that cannot go down at Slack, Block, Etsy, Intercom, Cursor, MyFitnessPal, and Barstool Sports. Clusters fail over automatically across three availability zones on AWS and Google Cloud. The company is building Neki, a Postgres sharding layer built on Vitess lessons. That context explains why the current open roles (Software Engineer for Neki Orchestration, Insights, Sharded Postgres, PlanetScale Vitess (EMEA), and PlanetScale Postgres) all carry the same salary band and demand demonstrated database-engine experience.

Six in seven listings show pay; the band is global, not geo-adjusted. An EMEA Vitess engineer sees the same $120k–$290k range as a Bay Area peer. That parity draws senior talent in lower-cost regions but floods the funnel — so the technical screen filters harder on Vitess internals and distributed-systems judgment.

Cultural fit is explicit, not aspirational. The mission statement names specific protected classes and promises proactive interview accommodations. The brand-designer posting spells out concrete behaviors: juggle multiple deadlines without a manager chasing you, take feedback without surrendering the concept, write and speak clearly enough to defend choices and absorb direction. Designers also need curiosity about the domain, not just the craft; it's a signal that domain curiosity crosses functions.

Remote as Constraint, Not Perk

Remote-first is a constraint that dictates who can be hired and how. The company operates with three in four of its ~125-person workforce fully remote, one in four on-site (San Francisco heavy), zero hybrid. That distribution requires the process to handle time-zone overlap, legal employment frameworks, and communication patterns a co-located team ignores.

Each opening carries a location tag ("Remote - United States," "Remote - Americas," or "EMEA - Remote") that decides which employment laws apply before a résumé is read. Sanctions screening is the first gate. Embargoed countries (North Korea, Cuba, Iran, Syria, Sudan, Crimea) are barred; Mainland China is excluded for separate legal and practical reasons; country-specific sanctions lists block individuals flagged for human-rights violations, terrorism, narcotics trafficking, or genocide. This vetting happens before any technical assessment.

The company hires in 24 countries, but the volume clusters: four listings in the U.S., three each in Germany, Egypt, South Africa, Serbia, Spain, France, the United Kingdom, Romania, Ireland, Portugal, Italy, Kenya, Nigeria, and the Netherlands. The "24 countries" figure shows where hiring is legal, not where it happens at scale. The U.S. and Western Europe see deeper applicant pools; candidates in other regions pass compliance only to find few teammates in their time zone.

Employment model adds another variable. Direct hire where PlanetScale has an entity; PEO elsewhere. Direct hires get PlanetScale equity and benefits; PEO hires get local contracts and a different equity structure. The posting doesn't always say which model applies; recruiters clarify in the first call. That conversation also tests the candidate — those who ask about entity structure, tax treatment, or equity vesting show the systems thinking the bar requires.

Role tags imply time-zone expectations. An EMEA role expects CET overlap; an Americas role expects Pacific-to-Eastern overlap. The process tests this without a separate culture-fit interview. Candidates who write clearly, version their thinking, and resolve ambiguity in text move forward; those who need a whiteboard or verbal nuance stall.

Listings stay open about two months — long for a 125-person company. That duration reflects the compound filter: compliance, time zone, employment model, then technical depth. Each layer drops candidates who would pass a standard remote screen but fail PlanetScale's constraints. The pipeline looks slower but yields engineers who already work the way the team works.

Inside the Loop

PlanetScale uses Greenhouse, the same ATS used by Airbnb, Slack, HubSpot, and Dropbox. The platform is just plumbing; what matters is configuration. PlanetScale configures it for depth, not speed.

A human, not an algorithm, reviews every application. Greenhouse doesn't auto-reject on keyword density. A recruiter scores each resume against "Focus Attributes" defined for the role, attributes lifted straight from the job description. Each bullet maps to a scorecard row. If your resume doesn't give concrete evidence for a bullet, that box stays empty. At high-volume companies recruiters spend under ten seconds scanning; at PlanetScale, with a small senior engineering team, the review is more deliberate — but the principle holds: your most relevant experience must be unmissable in the top third of your resume.

Candidates who pass enter a process that typically runs five stages, varying by role. Most funnels have two to four live stages after the screen: a recorded video or phone screen, one or more competency interviews, and a final panel. In the U.S. the typical sequence is recruiter screen, hiring-manager loop, final panel. Mid-level loops close in two to six weeks; senior loops run longer when panels and compensation approvals stack.

The recorded video screen is a PlanetScale hallmark. Many roles ask candidates to answer timed questions on camera with no live interviewer. It shortlists candidates before live rounds. The questions are standard: tell me about yourself, why PlanetScale, describe a relevant project, salary expectations, timeline. But the format penalizes vague answers. Generic responses like "I always try to communicate clearly" score a 1 or 2 because there's nothing specific to rate. Candidates who treat it as a warmup rarely advance.

Every live interviewer uses the same Interview Kit: four to eight fixed questions per role. This is deliberate; it makes scoring comparable. Greenhouse trains interviewers to score defined attributes; PlanetScale follows suit. Behavioral rounds probe ownership, judgment, and autonomy. The company hires senior engineers who work autonomously. A candidate who can't describe a conflict resolved, a trade-off owned, or a project driven from ambiguity to production stalls here.

Technical rounds decide the outcome. Most candidates face coding, system design, and a database-systems deep dive. Coding rounds favor practical problems over algorithmic tricks: string manipulation, heaps, rate limiters, schedulers. System design covers database-platform prompts: sharding strategy, consistency guarantees, online schema changes, multi-cloud failover. The PlanetScale-specific deep dive is the differentiator: be ready to reason from first principles about database internals, including Vitess architecture, sharding strategies, non-blocking schema changes, replication topologies, and query planning. The database and systems rounds are where the bar lives.

The final panel includes the hiring manager and bar. Surprises at this stage (notice period, relocation, competing offers) slow or kill strong loops. Greenhouse launched an anti-ghosting initiative in late 2024; adoption varies. At PlanetScale, "Active" status can mean the rejection is queued. No response in three weeks often means the role is inactive. A referral doesn't guarantee an interview but gets your resume in front of a recruiter faster with a positive signal. Silence after a strong round usually means scheduling; a polite nudge after ten business days is fair. After that, move on mentally.

What Separates the Candidates Who Advance

The 13 open roles have pulled a wide applicant pool. Engineering base pay runs $120k–$290k, Zero G Talent's board found (median $250k); total compensation from reported offers spans $40,000 to nearly $1 million depending on level and location. PlanetScale competes directly for talent with OpenAI, ByteDance, Instacart, Affirm, Amazon Services, and Uber, companies also hiring for database-platform and distributed-systems depth.

The loop rewards specificity. A dozen reported interviews: five in six rated medium difficulty, one in six easy, none hard; five in six positive sentiment. The content breakdown is the signal:

Candidates who substitute "I've used Postgres" for MySQL internals knowledge fail. So do those who hand-wave replication mode, sharding-key selection, or cross-shard query resolution. The branching feature (non-blocking schema changes with a git-like workflow) is a differentiator. Interviewers expect candidates to explain the coordination complexity, not dismiss it as simple.

Process mechanics matter. Most loops run two to four stages, often opening with a recorded timed video screen before live rounds. Large-employer loops punish silence more than imperfect answers. Prep guides recommend: map the funnel, drill timed skills, write STAR stories with metrics, mock interview, debrief. Skipping the debrief repeats the mistake. Candidates lose points for generic resumes, skipping behavioral prep, or showing up without notice-period and location constraints ready.

The remote-first model (hubs in San Francisco, Los Angeles, New York, plus global remote) expands the talent pool but raises the communication bar. Interviewers at a ~200-person senior engineering org hear dozens of "we shipped X" answers. The winning answer names team size, constraint, stakeholder, owned outcome, two to four concrete steps, and a metric. Memorizing funding history or founder names substitutes poorly for solving sharding, replication, and zero-downtime migration problems out loud.

Market-wide, the surge signals that database-platform specialization commands a premium. Supabase and Neon bet on Postgres; Aurora offers managed MySQL and Postgres on AWS. PlanetScale bets on MySQL/Vitess (a decade of YouTube production) with multi-cloud and enterprise compliance (key management, audit) driving the roadmap. Candidates should choose based on which technology foundation they want to deepen, not brand. Compensation is comparable across peers; interview rigor is solid everywhere.

The takeaway: demonstrate MySQL internals fluency, speak Vitess architecture concretely, practice timed system design on sharding and replication, and tailor your narrative to the specific team: Neki Orchestration, Insights, or core Vitess.

The Vitess lineage that once scaled YouTube's MySQL now faces a different test: whether the same architecture can absorb the write patterns and latency demands of agent-driven workloads. The 13 roles are the company's answer — not a hiring plan, but a technical thesis stated in headcount.

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