19:I[4003,["/_next/static/chunks/07c9a3d5c5a6f6ad.js?dpl=dpl_7tXqf7SWKc8dfo5d2EonWbVhE1Dm","/_next/static/chunks/46c092985e1a2451.js?dpl=dpl_7tXqf7SWKc8dfo5d2EonWbVhE1Dm","/_next/static/chunks/5576ad631a3f1b7f.js?dpl=dpl_7tXqf7SWKc8dfo5d2EonWbVhE1Dm"],"default"] 1a:T4b2,

You build the applications customers actually use. Full-stack, at a pace that isn't normal. You're equally good debugging an agent loop at 2am and demoing the feature to an ops lead the next morning.

What you own. AI-native applications end-to-end - React + TypeScript on top, Python + FastAPI + Postgres underneath, LLM orchestration in the middle. Prototype → production → iteration. Evals, observability, and the unsexy plumbing that keeps the whole thing running. Direct customer contact when it's the fastest way to build the right thing.

What you bring. Production experience across React/TypeScript, Python, and at least one LLM API in a real application. Strong opinions about RAG, agents, and evals, formed by shipping. You translate between a stakeholder's problem and the system that solves it without needing a PM in the middle.

Level. Experienced engineers own feature areas end-to-end. Early-career ramps fast on live customer systems.

1b:T4cb,

You build the systems behind the systems. Every Ciridae customer deployment runs on the platform you maintain - the leverage of your work compounds with every new engagement. This is the role for engineers who would rather solve the hardest infrastructure problem once than the same problem six times.

What you own. The hard problems behind running production AI across many live customer environments at once. Multi-tenant infrastructure, reliability and cost controls for non-deterministic systems, evaluation and observability where the ground truth keeps moving, and the internal tooling that turns what we learn on the ground into leverage for the next deployment. The questions a seed-stage team would normally postpone - we answer them now, because we have to.

What you bring. Deep systems engineering - distributed systems, APIs, databases, cloud infra. Expert Python (Go or TypeScript a plus). Terraform, GCP, containers, multi-tenant production experience. Taste for the LLM-specific infra problems: model routing, evals, cost and latency controls.

1c:T52e,

This is a super-powered FDE who has world-class engineering talent with the ability to translate it to a business audience. Over four weeks, a mess of spreadsheets, disconnected SaaS, and tribal knowledge becomes a full-stack production system the business runs on. When the CFO calls at 10pm, you answer.

What you own. The full solution, end-to-end. Requirements, architecture, frontend, backend, data, LLM orchestration, production cutover, and the metric that tells us it worked. You're writing React in the morning, shaping a Postgres schema after lunch, and debugging an agent loop before dinner. Whatever you learn in the field feeds back into the platform.

What you bring. Production software shipped end-to-end across the stack — React and TypeScript on the frontend, Python on the backend, at least one LLM API in real use. Opinions about RAG, agents, and evals formed from building them, not reading about them. Comfort in a room with a CFO or a superintendent. Calm with ambiguity.

Level. Experienced FDEs lead engagements start to finish. Earlier-career engineers ramp quickly to do the same.

1d:T4ad,

Primary point