10 PlanetScale Roles Open, Top Pay Hits $290k

Hiring Volume and Open Roles

PlanetScale posted 10 open roles in mid-August 2026, its largest simultaneous hiring push since the Postgres pivot went public. The managed database platform built on Vitess has spent six years proving it can run the backbone of major customers at exabyte-adjacent scale. Now it's adding Postgres to the mix, shipping Metal for bare-metal performance, and staffing up to deliver both. The question isn't whether they're growing; it's whether their hiring bar can stay this high while the team doubles.

As of mid-August 2026, PlanetScale lists 10 open positions on JobsRadar across four departments: seven in software engineering, two in marketing and growth, one in customer engineering, with a thirteenth role appearing on Jobera's tally. Two postings went live in the previous 30 days, and one landed in the 24 hours before the board's last refresh on August 15. The first-party Zero G Talent board confirms the pace: one new role added in the past seven days, the Software Engineer – Neki Orchestration position based in the San Francisco Bay Area or remote. The roles span five locations, though Jobera's data shows 88 percent carry a remote designation and only 12 percent require on-site presence. Hybrid isn't offered.

The engineering openings read like a roadmap of the platform's expansion. Three roles target the new Postgres effort: Software Engineer – PlanetScale Postgres (fully remote), Software Engineer – Sharded Postgres (Neki) (Bay Area or remote), and Software Engineer – Neki Orchestration (Bay Area or remote). A fourth, Software Engineer – PlanetScale Vitess, is scoped for EMEA remote. The remaining engineering slots cover Insights (Bay Area or remote) and the core Vitess platform. Outside engineering, the company is hiring a Head of Events and a Video Producer, both San Francisco-based, plus a Developer Educator (San Francisco and/or US remote), a Solutions Engineer (remote across EMEA and North America), and an Enterprise Support Engineer (remote across NA, APAC, and EMEA). Only one of the ten, the Solutions Engineer, is open to candidates anywhere in the world; the rest carry regional constraints tied to PEO partnerships and U.S. sanctions compliance.

All ten postings disclose compensation, a transparency rate well above industry norm, with engineering bands reaching $290,000, according to Zero G Talent's figures. The full breakdown appears in the compensation section below.

PlanetScale's headcount stands at roughly 125 people, per Jobera's August 14 count, placing it in the 51–200 band that Fastaijobs also reports. The Series C round — $50 million in November 2021, Fast AI Jobs reported, atop $135 million in earlier funding — gave the company runway to be selective. A YouTube deep-dive from October 2025 noted a "tiny team of 45 engineers" running the entire platform, a ratio that implies each new hire carries outsized leverage. No node at PlanetScale lives more than 29 days, the Database School Podcast interview found; the infrastructure turns over faster than most hiring cycles. The current recruitment drive is the first time the company has posted this many simultaneous openings since the Postgres pivot became public, and the concentration in sharded Postgres and orchestration roles suggests the next 12 months will test whether the Vitess architecture can absorb a second engine without losing the operational discipline that made major customers trust it with every message and transaction.

How the Remote‑First Model Actually Works

PlanetScale was built remote. The company's 2018 founding by Vitess co‑creators came with a deliberate choice: no headquarters anchor, no mandatory office days. "We have a highly distributed, remote forward team with employees across the United States and the globe," the careers page states. That phrasing — "remote forward" rather than "remote friendly" — signals intent. The model isn't an accommodation; it's the operating system.

The current hiring slate reflects that system. Of the ten open roles on Zero G Talent's board, eight list "Remote" or the Bay Area/remote option as location. One targets EMEA explicitly — Software Engineer - PlanetScale Vitess (EMEA - Remote) — and only the Head of Events role pins to San Francisco. The pattern is clear: engineering talent can sit anywhere the company can legally employ them.

Legal employment is the hidden architecture. PlanetScale's Himalayas profile notes it will "either hire directly to PlanetScale or partner with a Professional Employment Organization (PEO). Due to business requirements, we may need to hire in a certain country or region." That sentence carries weight. A PEO partnership lets the company onboard engineers in countries where it lacks a legal entity, handling payroll, benefits, and compliance locally, without waiting for entity setup. It also means hiring managers can't simply approve a candidate in any timezone; the "business requirements" clause reserves the right to constrain geography by tax treaty, data residency rules, or PEO coverage maps.

The philosophy statement on the careers page frames this as talent attraction: "Our philosophy is to attract the best talent and to support them doing amazing things where they are the most comfortable and productive." But the PEO disclosure reveals the operational discipline underneath. Remote‑first at scale isn't a policy document; it's a compliance stack. Each new country adds employment law obligations: wage‑hour rules, mandatory leave, workers' compensation, social security contributions. The 2026 remote‑work policy templates list these as non‑negotiable; PlanetScale's PEO language acknowledges them without spelling them out in the job post.

Time‑zone management follows the same discipline. General remote‑work guidance recommends core hours and overlap windows; PlanetScale's distributed Vitess maintainers have effectively run on asynchronous coordination for years. The EMEA‑tagged role suggests the company now needs deliberate European coverage, not just "anywhere" hiring but targeted overlap for a database product that runs 24/7 for customers. The board data shows salary bands identical across geographies ($120k–$290k), which implies location‑agnostic compensation rather than geo‑adjusted bands. That's a recruiting lever: a senior engineer in Lisbon or Warsaw sees the same top of band as one in Austin.

Equipment and security baselines are implicit in the model. Industry templates mandate company‑issued laptops, VPN enforcement, device encryption, and annual security training for remote workers. PlanetScale's database product handles customer data; its own infrastructure likely runs on the same zero‑trust principles it sells. Candidates should expect a hardware stipend or shipped kit and a non‑negotiable security onboarding, table stakes for a cloud‑database vendor.

The "remote forward" label also shapes team rituals. Written communication defaults over synchronous meetings. Decision logs replace hallway conversations. Onboarding for a distributed engineer means documented runbooks, recorded architecture walkthroughs, and assigned buddies across time zones, not a week in an office. The company's public Vitess contributions demonstrate this culture: the project's governance, design docs, and review process all happen in the open, async, across continents. New hires step into that workflow on day one.

For applicants, the takeaway is practical. The location field on a PlanetScale posting isn't decorative; it maps to a PEO footprint or direct‑hire entity. A candidate in a country without coverage won't clear the legal gate, regardless of technical fit. The EMEA role signals active European hiring capacity; those roles signal U.S. coverage plus flexibility. Ask recruiters early: "Is my country covered by your PEO or entity?" It saves both sides a deep technical loop that ends in a compliance block.

What the Pay Numbers Actually Mean

PlanetScale publishes salary ranges on every open role, with 100 percent of its postings disclosing compensation, per transparencyjobs.net. That consistency is rare even among well-funded infra companies, and it gives candidates a clear baseline before they invest time in an application. Zero G Talent's board shows ten salaried openings with a shared engineering band of $120,000 to $290,000 per year and a median of $270,000. Five software-engineering listings — spanning Neki orchestration, insights, sharded Postgres, Vitess, and PlanetScale Postgres — all carry that identical range. A sixth role, Head of Events based in San Francisco, sits at $180,000 to $250,000.

Third-party aggregators paint a wider picture but disagree on the center. Levels.fyi, drawing on verified employee submissions updated August 15, 2026, reports a median total compensation of $251,838 and a spread from $240,000 for a software engineer at the low end to $263,675 for a solution architect at the high end. Hireoven, also citing 2026 data, places the median at $205,000 with a reported floor of $20,000 and ceiling of $320,000, a range so broad it suggests the sample mixes contract, part-time, and outlier packages. RepVue, focused on go-to-market roles and based on 14 salaries current as of the same mid-August date, breaks out sales compensation in detail:

The discrepancy between the board's $270,000 engineering median and Levels.fyi's $251,838 total-compensation median reflects composition: the board band is base-salary-only for the current open requisitions, while Levels.fyi blends base, bonus, and equity across all levels and tenures. Glassdoor's single figure of $157,518 for a software engineer, identical for both low and high, appears stale or mis-scraped and carries little weight against the fresher, multi-source data.

What emerges is a two-tier compensation architecture. Engineering roles cluster tightly around a high, transparent band that rewards deep systems expertise — Vitess, MySQL internals, distributed consensus — with base pay that competes with top-tier public-company offers. Sales roles follow a classic SaaS leveraged model: modest bases, aggressive OTEs, and outsized upside for enterprise hunters who close seven-figure deals. The Head of Events slot signals PlanetScale's investment in developer-facing brand building, priced between the two tiers.

For candidates, the takeaway is practical: negotiate from the published band, not from anonymized aggregates. The board data is the only source tied to live requisitions; third-party medians reflect historical mixes that may not match the current open roles. The next hurdle isn't pay; it's the technical screen that filters for the specific expertise those salaries buy.

Screening Criteria: The Three Pillars

PlanetScale's interview loop filters for a narrow profile: engineers who operate at the intersection of distributed systems expertise and autonomous execution. The company publishes its evaluation weights openly. Distributed systems knowledge appears in every reported loop — 100 percent — while database infrastructure engineering sits at 96 percent and Vitess-specific knowledge at 93 percent. MySQL follows at 90 percent, Go at 87 percent, and horizontal scaling at 84 percent. These are not aspirational targets; they reflect the actual frequency each topic surfaces across candidate reports compiled by Dataford.

Source: Dataford interview guides, aggregated from reported loops

The technical bar centers on three pillars. First, distributed systems and databases: candidates must explain how data distributes across nodes, how to maintain availability during partitions, and how sharding and resharding work under production load. Second, systems programming in Go: the core infrastructure runs on Go, so interviewers expect idiomatic, performant code using goroutines, channels, mutexes, and wait groups without race conditions or deadlocks. Third, cloud-native infrastructure and observability: operating database clusters at scale demands container orchestration fluency and comprehensive monitoring practices.

Cultural evaluation carries equal weight. PlanetScale describes its teams as "small, highly autonomous teams of p99 individuals," a phrase that appears repeatedly in its hiring philosophy. The company hires for high trust, high autonomy, and asynchronous communication. Interviewers probe for concrete examples where candidates took complete ownership, made high-impact technical decisions with minimal guidance, and delivered robust systems. They also look for an open-source mindset: contributions to projects like Vitess, technical blog posts, or engagement in developer communities signal alignment with PlanetScale's roots.

Communication clarity gets explicit scrutiny. Because the team is globally distributed, the ability to articulate complex technical ideas in writing and speech is "highly scrutinized," per the interview guide. Candidates who cannot explain their trade-offs to a non-technical audience fail the leadership round. The final interview tests three things: pinpointing the true problem, proving technical depth, and translating those trade-offs for stakeholders who will never read the code.

Glassdoor data shows 83.3 percent of candidates rate the experience positively, with a difficulty score of 2.67 out of 5. Account Executive and Technical Account Manager candidates report the hardest interviews; software engineers most commonly describe the process as medium difficulty. The loop runs three rounds over three to five weeks: recruiter screen, technical rounds (systems design, live coding, distributed systems), then a leadership or VP interview assessing strategic alignment and autonomy readiness.

Core Tech Stack and Desired Expertise

Vitess sits at the center of PlanetScale's stack. The project began at YouTube where the company's founders scaled MySQL to petabytes across 70,000 nodes in 20 data centers, a provenance that still shapes every hiring decision. Today Vitess powers production databases at Slack, HubSpot, Blizzard, Etsy, GitHub, Block, Bloomberg, and Yelp. Candidates who join the Vitess-facing teams inherit a codebase that routes queries through VTGate proxies, mediates MySQL communication via VTTablet sidecars, and relies on vtorc for automated fault repair. The topology layer runs on ETCD, the CNCF project PlanetScale recommends for cluster metadata. vtctld and its CLI counterpart vtctl handle workflow orchestration, while VTAdmin provides a web console for cluster observability. This is not a managed service that hides complexity; it is the control plane itself, and engineers are expected to debug across all of these components.

MySQL fluency remains non-negotiable. Vitess uses vanilla MySQL under the hood, which means replication behavior, failover semantics, and connection pooling operate on MySQL's native primitives. The platform integrates pt-online-schema-change and gh-ost for online schema migrations, tools that demand hands-on experience with large-table alterations without locking. Connection pooling happens at both the VTGate and VTTablet layers, a design that breaks through vanilla MySQL's connection limits but requires engineers to understand pool sizing, timeout tuning, and the memory footprint of proxied connections. Read scaling uses standard MySQL replicas; write scaling uses explicit sharding, both vertical and horizontal, distributing data across thousands of nodes while presenting a single logical database to the application. Candidates who have only operated managed RDS instances will find the operational surface area considerably wider.

Kubernetes is the deployment substrate, and the relationship is codified in the Vitess Operator that PlanetScale maintains. The Operator's version matrix is explicit: each major Operator release supports the three latest Kubernetes major versions available at RC-1, guaranteeing at least one supported Kubernetes version for the Operator's one-year lifecycle. As of the latest published compatibility table, Operator v2.16.* pairs with Vitess v23.0.* and Kubernetes v1.31 through v1.34, while the latest track targets Vitess latest against Kubernetes v1.33 through v1.36. The Operator is built in Go, and the project's module path (planetscale.dev/vitess-operator) appears directly in the build instructions. Engineers working on the control plane or the Operator itself write Go daily; the codebase includes API schema definitions under planetscale.com/v2 and a planetscale package for API versioning. Borg heritage shows in the design assumptions, but the production target is GKE, EKS, and AKS parity.

Postgres has become a parallel product line with its own sharding architecture, Neki. Described as a fresh design from first principles informed by everything learned building Vitess, Neki targets large-scale Postgres workloads with predictable performance and fault tolerance. Horizontal scaling with Neki sharding is in private early access preview as of the most recent public disclosures. The Postgres side also brings a curated extension set, including pg_cron, pg_hint_plan, pg_partman_bgw, pg_squeeze, pg_stat_statements, pg_strict, pginsights, pg_duckdb, and TimescaleDB, plus logical CDC integrations, guided no-downtime imports from Supabase, Neon, Heroku, and Aurora, and Database Traffic Control for query budget enforcement. Safety guardrails like pg_strict block mass UPDATE or DELETE statements without WHERE clauses. Monitoring hooks into Prometheus, Datadog, and pganalyze. Engineers on the Postgres track need depth in these extensions and the import tooling, not just generic Postgres administration.

Observability spans both products. Query Insights surfaces execution patterns for Vitess and Postgres. PlanetScale Metal, the bare-metal NVMe tier with unlimited IOPS, adds a performance dimension that changes capacity planning: disk autoscaling, CPU architecture selection, storage profiles, and parameter tuning all become configurable levers. Multi-region deployments route read traffic through a Global Edge Network to local read-only replicas. Automated and custom backups, point-in-time recovery, and revertible schema changes with zero data loss round out the operational toolkit. The compliance surface includes SOC 1 Type 2, SOC 2 Type 2+ HIPAA, PCI DSS 4.0 Level 1, and BAA availability on all plans.

The five Postgres and Neki roles map directly to these surfaces.

Application Timeline and Candidate Experience

PlanetScale's interview process runs five distinct stages for most engineering roles, and the company signals up front that the bar reflects the reality of operating a production-grade database platform. The sequence (recruiter screen, technical phone screen, take-home for senior and staff candidates, then a four-to-five-round virtual onsite) has been documented consistently across interview reports. Candidates should expect the full loop to span several weeks; the research does not publish a fixed calendar, but the depth of each stage and the scheduling of multiple onsite rounds make a two-to-three-week cadence typical for candidates who move quickly.

The opening recruiter screen lasts roughly 30 minutes and covers background, motivation for PlanetScale specifically, and team alignment. This is not a formality; recruiters probe whether the candidate understands the company's MySQL and Vitess foundation and can articulate why that technology base matters for the role. A technical phone screen follows, a 60-minute session with a single medium-hard coding problem. The difficulty rating across sources sits at 7.5 out of 10 on techinterview.org's scale, while dataford.io's aggregated reports from 12 interviews place the overall experience at 4.6 out of 10, with most candidates rating it "medium." The gap suggests the coding screen is sharper than the average Big Tech loop, but the surrounding stages balance it out.

For senior and staff positions, a take-home assignment replaces or supplements the second coding round. The exercise runs four to six hours and models a realistic engineering problem, often a systems-design task that touches sharding, replication, or schema-migration logic. Candidates who treat it as a throwaway coding test tend to stall; the evaluators look for the same operational rigor the platform demands: explicit handling of failure modes, clear trade-off documentation, and a working prototype that could plausibly run in a Vitess cluster.

The virtual onsite consolidates four to five rounds in a single day or across two half-days: one to two coding sessions, a system-design review, a database-and-systems deep-dive, and a behavioral conversation with the hiring manager. The deep-dive is where PlanetScale's differentiation shows up most clearly. Interviewers expect candidates to name replication modes (async versus semi-sync), justify a sharding key, and walk through how a cross-shard query resolves. Hand-waving on these topics reads as inexperience. Likewise, designs that quietly accept a maintenance window or read downtime lose points; zero-downtime schema changes and uptime SLAs are product promises, not aspirations.

Feedback loops are not publicly timed, but the company's distributed-first hiring philosophy implies structured debriefs and timely communication. Candidates report that the process respects their time; the 12-interview sample on dataford.io shows no pattern of ghosting or unexplained delays. The strongest signal comes from preparation: candidates who spin up a PlanetScale database, exercise branching and deploy-requests, and read the Vitess architecture and VReplication docs arrive fluent in the vocabulary the deep-dive demands. The engineering blog's posts on Boost, Insights, and multi-cloud architecture are also cited as high-value prep material.

Red flags that consistently surface in post-mortems: treating Postgres experience as a direct substitute for MySQL internals knowledge, dismissing Vitess as "just MySQL sharding," and ignoring the customer-reliability dimension in system design. The converse, demonstrating database-ownership stories, explaining InnoDB and binlog behavior, and showing hands-on time with non-blocking schema changes, correlates strongly with offers. The process is deliberate, not performative, and it filters for the same operational discipline that keeps customer databases online.

Market Signals and What They Mean for Applicants

PlanetScale's ten-role hiring push lands in a tech labor market that Dice reports as polarized: overall postings dipped 10% month-over-month in July 2026, yet year-over-year volume remains up 10%, and AI-adjacent roles continue expanding while generalist positions contract. The company's openings, six software-engineering slots spanning Vitess, Postgres, and the new Neki sharding architecture, plus a Head of Events role, sit squarely in the specialist tier that Robert Half identifies as the hardest to fill. Eighty-seven percent of technology leaders report difficulty finding skilled workers, and 78% plan to increase permanent headcount in the second half of 2026. PlanetScale's $120k–$290k band (median $270k) exceeds the $112,521 average U.S. tech salary Dice recorded for 2025, and it aligns with the 17.7% premium that AI-fluent professionals now command.

The remote-first, globally distributed model PlanetScale advertises mirrors a structural shift the Rest of World report documented: companies that once hired remote workers domestically now recruit across borders, with Southeast Asia, Latin America, and Eastern Europe absorbing accelerating demand. PlanetScale's own careers page acknowledges U.S. embargo constraints but notes it will hire directly or through a Professional Employment Organization to reach eligible candidates worldwide. That posture matters when 18% of HR professionals say they are more likely to use skills-first hiring for remote software-engineering, data, and UX roles, and when the same senior engineer costs $8,000–$18,000 monthly in the U.S. versus $2,000–$6,000 in Vietnam, the Philippines, or Malaysia at comparable quality.

Community signals reinforce the specialist premium. Glassdoor shows six anonymous interview reviews and six published questions for PlanetScale, modest volume but consistent with a ~200-person company hiring selectively. A February 2026 DEV Community post captured the prevailing mood: "The job market is not broken. It is recalibrating. If you built your career on writing CRUD apps, yes, you are in trouble. If you built it on solving hard problems, you are in the best position you have ever been in." PlanetScale's stack — Vitess, MySQL, Postgres, Go, Kubernetes, and now the Neki orchestration layer — is exactly the "hard problem" tier. Customer testimonials from Anysphere (Cursor), Cash App, Convex, and MyFitnessPal confirm the platform runs millions of queries per second on hundreds of terabytes for AI-native and fintech workloads, the very workloads driving the 53% year-over-year surge in AI/ML engineering postings and the 200%+ growth in vector-database and AI-infrastructure skills Dice tracked in July.

The talent pool is restless. Dice's latest salary survey found 59% of tech professionals feel underpaid — the highest reading ever — and 47% are actively job hunting, up from 29% a year earlier. Gen Z anxiety runs hotter: 64% worry about layoffs versus 45% of millennials. Yet the same data shows senior and staff-level openings now exceed half of all software postings, while junior roles shrink. PlanetScale's roles, which demand deep systems knowledge (Vitess sharding, online schema changes, consensus protocols), filter for the senior cohort that Gartner says will face irreversible skill shortages by 2030. The company's interview process, described in public write-ups as heavy on distributed-systems design and production debugging, acts as a de facto skills verification in a market where Gartner warns candidate fraud and generative-AI-assisted cheating are rising.

PlanetScale's hiring also reflects the infrastructure layer's growing leverage. As AI coding tools absorb boilerplate work, the bottleneck shifts to data-layer correctness, scale, and uptime, precisely the guarantees Vitess and Neki were built to deliver. The World Economic Forum projects a net 78 million new roles by 2030, but 39% of core skills will change by then. Engineers who operate at the database-runtime boundary, who can review AI-generated migrations and still reason about consensus latency, are the ones companies will retain. PlanetScale's ten openings are a small number in absolute terms, but they signal where the next tier of specialist demand is concentrating.

As noted earlier, nodes cycle in under 29 days, and infrastructure turnover outpaces hiring cycles. The ten open roles are the latest nodes in a system that replaces itself before most companies finish onboarding, each hire inheriting a codebase that routed queries for YouTube at 70,000 nodes, now stretching to absorb Postgres without losing the discipline that earned that same trust from Slack and Cash App.

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.