Who gets hired and onto which teams
Of the eight roles Mindpeak has open as of March 2026, three sit in People and HR: a People and Business Operations Specialist, a Senior Talent Acquisition Manager, and a Working Student Office Management Support. That ratio — nearly half the hiring slate dedicated to building the machinery of hiring itself — signals a company that has de-risked its core AI pathology technology and is now assembling the organization to deliver it at scale.
Mindpeak employs about 60 people worldwide, up from 38 three years ago, a gain of more than half that hasn't followed a straight line. Headcount dipped to 37 in late 2023 before recovering, Revelio Labs data shows. Nearly all staff sit in Western Europe; the rest split between Eastern Europe and North America. Every open role is based in Hamburg, and no one works remote.
The functional split tells the story. Finance and Operations makes up nearly two-fifths of the workforce and is growing fastest, up about a fifth year over year. Engineering holds about a third. Sales and Marketing just under a third. Engineering roles remain targeted: the board lists a Full-Stack Software Engineer (React, TypeScript, Python/Django), a Head of Engineering, and a (Senior) Full-stack Developer - Backend Focus (Python). The stack (Python, Django, FastAPI, React, PostgreSQL, Docker, Kubernetes) signals a production-grade web product, not a prototype. A VP Product – AI for Precision Medicine and a Senior Medical Project Manager round out the slate, bridging clinical validation and product direction.
Seniority skews experienced: half the open roles carry "senior" or "head/VP" titles. Hiring has slowed to about three roles a month in 2026, down from five in 2025, and the median time to fill a role stretches to four months. Fewer than one in a hundred applicants make it through, the careers page states. Founders Felix Faber and Tobias Lang set a deliberate filter: "We are deliberate about every hire," the site reads. "Our mission and the complexity of the problems we work on require exceptional people."
What it pays
The Zero G Talent board tracks live postings and derives a company salary band from roles listed on the platform. Its latest ingest shows a band of $21k–$31k with a median of $31k, based on a single salaried role. That one data point is all the board has; the other seven open postings disclose no salary. Only the VP Product – AI for Precision Medicine lists a rate: $10–15 an hour, inconsistent with a VP salary and likely signaling a fractional or mis-categorized posting.
The postings span a wide seniority range. The the working student role — a part-time, on-site role in Hamburg — typically pays German working-student rates of €12–16 an hour before taxes, though the board doesn't confirm it. Mid-level People roles like the People and Business Operations Specialist and Senior Talent Acquisition Manager typically command €50k–70k and €65k–85k in Hamburg health-tech startups, but the board has no first-party figures.
Engineering roles dominate the senior end. The Full-Stack Software Engineer and the backend-focused full-stack role sit in the mid-to-senior bracket. The Head of Engineering and VP Product – AI for Precision Medicine are leadership slots. The Senior Medical Project Manager rounds out the clinical-operations track. The board's only salaried data point — $31k median — is an outlier far below market rates for these roles.
| Role | Typical Hamburg Base (Series A–B Med-Tech) |
|---|---|
| Full-Stack Engineer (React/TypeScript, Python/Django) | €65k–90k |
| People & Business Operations Specialist | €50k–70k |
| Senior Talent Acquisition Manager | €65k–85k |
| Senior Medical Project Manager | €60k–80k |
The gap between the board's band and market rates usually means the captured role is junior or non-technical, the postings are equity-heavy, or the sample is too thin. Candidates should treat each posted role as the primary signal; ask recruiters for the band attached to the requisition, and use the board's median only as a floor.
How the hiring process works and what gets candidates through it
Mindpeak's five-stage process evaluates mutual fit at each step, moving from broad alignment to technical demonstration to cultural and mission fit.
Stage one: a 30-minute Google Meet with HR. The recruiter pitches the company and role while learning the candidate's background and motivation. It's a structured screen, not a casual chat; candidates who can't articulate why Mindpeak's mission in computational pathology matters to them rarely advance.
Stage two: a 20- to 30-minute Google Meet with the team lead. The conversation centers on experience, problem-solving, and how the candidate could contribute. Engineers discuss architecture decisions, debugging workflows, and stack experience. Operations and people candidates discuss process design, stakeholder management, and scaling practices.
Stage three: a skill assessment, typically a live coding challenge or practical exercise lasting about an hour. The company calls it a chance to "demonstrate your skills and thinking process in a real-world context." Candidates say the exercises mirror actual team problems (data pipeline construction, model evaluation, API design), not abstract puzzles. Evaluators prize clarity of thought, iterative refinement, and communication, not just a working solution.
Stage four: an onsite "get to know the team" session in Hamburg. Not a panel interview, but a series of informal conversations with future colleagues to assess day-to-day collaboration. The company frames it as two-way: candidates learn how the team works, the team gauges fit.
Stage five: a short conversation with CEO Felix Faber on mission, values, and long-term vision. Hiring authority rests with leadership, not just managers, and cultural coherence is a strategic priority.
Glassdoor reviews (six questions, three reviews) corroborate the structure and note behavioral and situational questions probing past work and critical thinking. The company says the process aims to "explore how you think, solve problems, and collaborate while giving you a clear view of our work, culture, and expectations." Candidates who treat each stage as a dialogue, asking substantive questions about pathology, regulation, and engineering, advance. Those optimizing for "right answers" over transparent reasoning don't.
Mindpeak accepts open applications from exceptional candidates when no role matches, though the same multi-stage process applies. The company doesn't offer internships but occasionally hires working students, such as the the working student role, who follow a condensed version of the same framework.
Where the work happens
Mindpeak's headquarters sits at Zirkusweg 2 in Hamburg's Hammerbrook district, a short walk from the central station and the UKE medical center biotech cluster. The address appears on every regulatory filing, partner agreement, and job posting; here lives the ISO 13485 quality system, 11 CE-IVD products, and the penetration-tested infrastructure serving pharma partners and diagnostic labs. The Cambridge, Massachusetts office handles East Coast commercial and scientific engagement, but the regulatory core, engineering team, and clinical operations keeping Mindpeak's AI compliant with European medical-device law are anchored in Hamburg.
The work dictates the space. Building AI that runs on multiplex immunofluorescence, brightfield IHC, and H&E slides, integrating with ZEISS Axioscan 7, SlideStream, RareCyte Orion, Akoya PhenoImager HT, and Lunaphore COMET platforms, means the office can't be a laptop farm. The facility houses compute heavy enough to train and serve the peakFrontier foundation model, storage meeting GDPR and clinical-trial retention rules, and a staging environment where on-premise dockerized deployments are validated before shipping to a partner's firewall. ISO 13485 isn't a plaque; it's a documented quality system touching code review, release gating, supplier audits of imaging hardware, and post-market surveillance, all requiring controlled meeting rooms, audit-ready documentation spaces, and version-controlled hardware benches for scanner integration testing.
Hybrid is the stated model; every open role lists Hamburg, but the careers page describes flexibility. In practice, the rhythm follows regulated development. Engineers on the PhenoScout viewer or peakAcademy backend can push code from home. But when a new mIF model needs validation against slides from the Institut für Hämatopathologie Hamburg or Charité Universitätsmedizin Berlin, someone must be in the room where the scanner sits, the LIMS integration is wired, and ground-truth annotations are reviewed with a pathologist. The same for regulatory teams preparing CE-IVD submissions, or medical project managers coordinating multi-site studies with Intermountain Health and Bio Reference Labs. The office is where the physical and digital product meet.
The team is small (11 to 50 people by LinkedIn), so the footprint is modest compared to a traditional medtech manufacturer. No cleanroom, no wet lab, no production line. Instead: a server room with the uptime and audit trails a QMS demands, collaboration spaces sized for cross-functional sprint reviews with clinicians and regulatory affairs, quiet zones for the deep concentration that model architecture and validation protocols require, and a front door that receives auditors, pharma partners, and scanner vendors without improvisation. The House of AI Hamburg, where CEO Felix Faber has spoken on moving AI from research into pathology workflows, is a short transit ride away, close enough that the local ecosystem is usable, not theoretical.
Travel is constant but headquarters is the fixed point. The LinkedIn feed shows the team cycling through World CDx in Boston, IAP in Fukuoka, World ADC and B4Tour in San Francisco, PathVisions in San Diego, but each trip originates from and returns to Zirkusweg 2. Conferences are where partnerships are signed and biomarkers debated; the office is where those commitments are engineered into software that passes an ISO audit. That distinction — between transient business development and permanent regulated product — is what the Hamburg facility supports.
Who thrives here
Mindpeak's careers page names three traits that bind its team: curiosity, ownership, and the ambition to build AI that truly works in healthcare. Not "passion for AI" or "drive to innovate," but a specific focus on systems that function in clinical practice. That distinction shapes who succeeds. The company sits at the intersection of machine learning, pathology, and regulated medical-device development. A candidate who optimizes model metrics on a benchmark but can't explain how a false negative changes a patient's treatment plan will struggle.
The team composition reinforces the profile. With 43 employees in Hamburg, Mindpeak employs scientists, engineers, physicians, and builders across disciplines and nationalities. Daily work requires translating between clinical workflows, regulatory constraints, and software architecture. A full-stack engineer might pair with a pathologist to define annotation protocols one day and refactor a Django service for audit-trail compliance the next. The Senior Medical Project Manager and the VP Product role roles signal that product decisions carry clinical and regulatory weight: ownership means carrying a feature from concept through CE marking.
Glassdoor reviews for a similarly named entity (Mindpearl) show three in four employees would recommend the workplace, with culture and career opportunities rated mid-range and work-life balance slightly lower. The naming discrepancy prevents direct attribution, but the pattern aligns with a company investing heavily in onboarding and cohesion. Candidates who treat the interview as a two-way assessment, asking how clinical validation shapes sprint planning or how the Head of Engineering balances technical debt against regulatory deadlines, signal the curiosity the company values.
Founders Faber and Lang built the company around a clinical problem: pathologist workloads are rising while specialist numbers stay flat. The tools Mindpeak ships (automation for cell quantification, biomarker scoring, reproducible diagnosis support) are used in partner pathology and university labs. That reality filters for people who tolerate ambiguity in model performance when the cost of error is a misdiagnosis, and who document decisions not for process compliance but for the next engineer who must maintain the system under audit. The board's salary band reflects early-career and working-student roles; senior hires negotiate individually. But the signal is consistent: Mindpeak hires builders who want their code to touch a slide scanner in a hospital, not a demo dashboard at a conference.
Working in frontier tech? Zero G Talent tracks the openings: see every open Mindpeak role, browse frontier tech jobs, the companies hiring, and the people building the field.