job specification

Optiak is the modular operating system for enterprise AI. We extract the platform logic companies normally scatter across applications integrations, routing, security, memory, observability, prompt enrichment, and optimization into seven standardized layers that work with any model and any third-party tool, under one governed framework.

Those seven layers are the product. The thing they all sit on - multi-tenancy, data isolation, secrets, deployment, observability backbone, the infrastructure under all of it - is what this role owns.

Our customers are large, regulated enterprises with strong opinions about where their data lives. Optiak ships to all of them - SaaS, private cloud, customer VPC, and on-prem. We need a founding engineer who can build the platform underneath the product and own how it runs in every one of those environments.

This is a deeply internal role. You won’t be in front of customers and you won’t be in pre-sales. Your customers are the engineers building Optiak, and your work product is the platform every other layer depends on.

WHAT YOU'LL OWN

Backend platform & core architecture

• Design and build the foundational backend systems every Optiak layer is built on top of - multi-tenancy, data isolation, identity, permissions, the internal APIs and service contracts other engineers depend on.
• Own the platform primitives the product needs from day one: tenant routing, configuration, secret management, audit trails, the connective tissue that makes seven layers feel like one product.
• Write production Python for the platform services this role demands - this is a hands-on engineering role, not an architecture-only one.

Internal frameworks & developer platform

• Build the internal frameworks and tooling other engineers at Optiak use to ship: service scaffolds, CI/CD, build and test infrastructure, preview environments, observability defaults.
• Make the platform’s defaults the secure, multi-tenant, observable ones - so every new service

 starts in the right place without anyone having to ask.

Cloud & multi-environment deployment

• You’ll own how Optiak gets into any environment we sell into - SaaS multi-tenant, single-tenant private cloud, customer-managed VPC, and on-prem.
• Design the deployment model: one product, multiple delivery modes, the same product surface in each.
• Build the packaging, release, and upgrade machinery that makes shipping to a regulated customer environment as routine as shipping to our own cloud.
• Own the platform’s behavior in environments you don’t control: air-gapped installs, customer-managed Kubernetes, restricted egress, and the long tail of “their security team says no.”
• Design for multi-region from day one, including EU data residency for customers who require it.

Reliability, observability & security primitives

• Define and own SLOs for the platform, the incident process behind them, and the on-call practices that keep them honest.
• Build the observability backbone — logs, metrics, traces — that the product’s Observability Layer runs on top of.
• Build the infrastructure primitives the Security Layer depends on: KMS/HSM-backed secrets, BYOK plumbing, RBAC, tenant isolation, the controls behind SOC 2, ISO, and GDPR posture.

WHAT WE ARE LOOKING FOR

Core

• 8+ years in backend, platform, or infrastructure engineering, with deep production experience building and operating systems for large customers.
• Genuinely hands-on. You write production Python today and want to keep doing so — this role is one part architecture, two parts code.
• Strong backend fundamentals: distributed systems, APIs, data stores, networking, identity, the practical tradeoffs across them.
• Strong Kubernetes and cloud background (GCP / AWS / Azure). Not "we run k8s," but you've designed against its sharper edges and operated it at scale.
• Comfortable across the modern infra stack: IaC (Terraform), observability tooling, secrets and identity, CI/CD.

 On-prem / customer-environment experience (this matters here)

• You've shipped a product into environments you don't operate — private cloud, customer VPC, air-gapped, or on-prem — and lived with the consequences.
• You understand what "compliance posture" actually means in practice: data residency, secret management, audit trails, the controls behind the auditor's checklist.

 Judgment

• You think in systems and platforms, not scripts. You've built things meant to be operated for years by people who weren't in the room when they were designed.
• You make architectural calls under real constraints — cost, blast radius, regulatory, deadline — and own the operational consequences of those calls.
• You're allergic to magic. The platform you ship should be debuggable by the next engineer at 3am with no context.

 AI-aware (nice to have)

• Familiarity with the infra patterns specific to AI workloads: model serving, GPU scheduling, retrieval/vector infra, the cost shape of LLM traffic.
• Experience supporting AI products in production environments.

WHY THIS ROLE

You'd be a founding member of a 14-person team building a category-defining platform, with direct ownership of the entire platform-and-infrastructure surface — the backend foundations, the deployment story, every environment we ship into. We offer a competitive salary and meaningful founding-level equity. You'll work next to a founding team from Google DeepMind, Amazon, Stripe, and Hugging Face, and shape how the next wave of enterprise AI actually runs in production.