Senior Platform Engineer

Talonic GmbH; Berlin; Vor Ort; EUR 80.000 - 90.000; Vollzeit

Benefits dieser Stelle

VSOP equity

Hybrid work in Berlin

Zusammenfassung

Talonic GmbH in Berlin is seeking a dedicated reliability leader to own the platform’s uptime, performance, and correctness as the first dedicated hire. You will define SLOs/SLIs, implement observability, lead incident response, and scale capacity for high‑volume document pipelines across energy, logistics, and procurement customers.

This role offers autonomy, a clear charter, and equity. Join a small, high‑trust team and help shape a reliability function expected to grow with the company.

Qualifikationen

• You’ve owned production systems relied on by real users.
• You remain calm and decisive under Sev1 incidents.
• You apply post‑mortem rigor and honest error budgets.

Aufgaben

• Define reliability with SLOs/SLIs and error budgets.
• Build observability, alerting, and on-call readiness.
• Lead incident response and blameless post‑mortems.
• Ensure safe rollouts and robust deployment pipelines.

Kenntnisse

Production ownership

Calm under Sev1

Post-mortem discipline

Jobbeschreibung

We run high-volume document pipelines for enterprises in energy, logistics, and procurement. In those environments a missed batch isn't an inconvenience; it’s a billing cycle or a compliance deadline.

As those commitments grow, reliability stops being something the founding team holds in its head and becomes a function with a single, named owner. You are that owner: the first dedicated hire whose mandate is that the platform stays up, stays fast, and stays correct under load. You'll start as an individual contributor with real autonomy, and build the reliability function the company grows into.

What you'll own

The charter

Define reliability

SLOs, SLIs, and an error budget the whole team deploys against. Turn “99.9%” from an aspiration into an instrument with a published policy.

See everything

Observability, alerting, and on-call that catch problems before a customer does, and make the system legible to everyone who builds on it.

Run incidents like an engineer, not a fire-fighter

Lead response under pressure, then run blameless post-mortems that change the architecture, not just the runbook.

Safe rollouts, fast rollback, a release pipeline nobody has to think about. Friday afternoons included.

Scale the throughput

Capacity planning and performance for high-volume ingestion and the API surface customers build on, engineered to stay flat as volume climbs.

Hold the line where reliability meets compliance

Work alongside our security and data‑protection posture (GDPR, ISO 27001 / 42001, HIPAA where it applies) so uptime and auditability move together.

First 180 days

What good looks like

DAY 30

Reliability, mapped

SLIs defined, current uptime measured honestly, the top failure modes named and ranked.

DAY 90

The budget is live

An error budget published and adopted. On-call and incident response exist and have been exercised for real. Deploy safety has a floor.

DAY 180

A function, not a heroics streak

99.9% is instrumented and trending. The reliability function has a shape other engineers can grow into.

Who you are

The profile

• You’ve owned production You’ve been on the hook for a real system real users depended on. You’ve been paged at 3am and made the right call. This is scar tissue, not a certification.
• You’re calm when it’s on fire Low ego, clear head, decisive under a Sev1. We screen hardest for this, because it’s the trait that can’t be taught.
• You’re disciplined and diligent Post-mortem rigour, honest error budgets, no cowboy deploys. Reliability is the craft of doing the unglamorous thing every single time.

What sets you apart

• You build agentically You don’t tolerate AI‑native engineering; you reach for it first. Our engineers ship with Claude Code. If that excites you rather than threatens you, we should talk.
• You’re ambitious about the boring stuff Your ambition points at making reliability a moat, the thing customers trust us for, rather than chasing the next greenfield. Mission over status.

The operating model

Small team, high trust, context over control. Real ownership from week one, and a weekly heartbeat everyone ships to.

Monday

Monday Business requirements set

Thursday

Thursday Delivery

Friday

Claude Code is how we ship, not an experiment. We build agentically by default, and we expect the reliability of an agentic codebase to be designed in, not bolted on.

We help write the rules of this category

• We co‑authored DIN SPEC 91491, the German standard for AI‑based document data extraction, with Fraunhofer IIS, Humboldt‑Innovation and DIN e.V. We don’t follow this category’s rules; we help define them.
• NVIDIA Inception member, with roots in research affiliated with Humboldt University of Berlin, and backed by leading European deep‑tech investors.
• Enterprise traction, at the inflection point where the platform underneath the promise has to be genuinely unbreakable. That’s the seat you’d be taking.

What we offer

• Location: Berlin, hybrid. CET plus/minus 2 for on‑call coverage.
• Compensation: Senior base of €80,000–€90,000 plus meaningful VSOP equity that reflects how early you’d be joining the team.
• Scope: The reliability charter for a company whose entire promise is permanence: yours to define, instrument, and grow.

How we’ll meet

The process

Intro with the founders. Meet the CEO and CTO: the mission, the seat, your questions. Bring a real Sev1 you’ve owned and walk us through it: the first ten minutes, the post‑mortem, and what changed structurally afterwards.

Paid co‑working session. Spend a paid day building alongside the team on a real problem from our roadmap, with our stack and the way we actually work.

Culture‑fit meeting with the team. Meet the people you’d build with. Mutual fit, no trick questions.

References. Including the ones you didn't put on the list.