Careers

What working here is like

Actonics builds production AI systems for mid-sized US companies: document and contract intake, claims processing, internal agents. The work is fixed-price and eval-driven, which shapes everything about the engineering culture. There's no timesheet theater and no demo-ware; a project is done when the accuracy report says it's done, on the client's real data. Everyone who builds also talks to clients. Nothing gets handed to juniors, because there are no juniors.

The team is distributed. Client-facing time lands in US business hours; deep work happens whenever you do it best.

Open role: Forward Deployed Engineer

You own pilots end-to-end. That means scoping one workflow with the client, getting their real documents flowing under NDA, building the evaluation set that defines "correct," shipping the extraction or routing pipeline against it, and writing the Week-4 accuracy report that drives a six-figure go/no-go decision. You present that report yourself.

What the work involves:

• Python, LLM APIs, OCR and document processing across messy real-world formats — scanned PDFs, fourteen-layout invoices, email threads.
• Designing eval sets and confidence-based human-review routing, then defending the numbers to the client's team.
• Integrating with the systems clients already run: CLMs, claims platforms, ERPs, shared inboxes with ancient APIs.
• Weekly client contact through a 30-day pilot, in US business hours.

What we need to see: production LLM or document-AI systems that real operations teams depend on daily, with war stories about the failure modes; the writing skill to make an accuracy report clear to a non-engineer; comfort being the technical face of an engagement.

Apply: info@actonics.com, subject "Forward Deployed Engineer"

Open role: Senior DevOps Engineer

Our systems ship into client environments, not ours — VPC and on-premise deployments where the client's data never leaves their perimeter. You make that repeatable, boring, and auditable.

What the work involves:

• Infrastructure as code (Terraform or equivalent) for per-client AWS/VPC and on-prem deployments.
• CI/CD for model pipelines, plus the monitoring that matters in this business: accuracy drift alerting, eval re-runs on model updates, inference cost tracking.
• Security posture a client's IT reviewer will sign off on: least-privilege access, audit trails, secrets handling, SOC 2-aligned practices.
• Being the person who makes "we deploy into your environment" a one-week fact instead of a slide.

What we need to see: deep AWS experience, Terraform in production, containerized deployments into environments you didn't control, and at least one data-sensitive deployment (healthcare, legal, insurance) where you dealt with a real security review.

Apply: info@actonics.com, subject "Senior DevOps Engineer"

How hiring works

One conversation about systems you've shipped, then a paid working session on a realistic problem from our domain — the same kind of work we do for clients, judged the same way. No leetcode, no take-home that eats your weekend, no five-round gauntlet. We know within a working session; so will you.

Not one of these roles?

If you're a staff-level engineer who fits the culture above but not these openings, write anyway — two or three paragraphs about a production system you shipped beats any resume. When the next role opens, the shortlist is where we look first, before we post anything anywhere. Same address, subject "Careers."