About Windward
Windward is the leading Maritime AI™ company, fusing data and Generative AI to empower global leaders with mission-grade intelligence. We help organizations anticipate threats and make high-stakes decisions with confidence.
This is a role for someone who wants to help build the org of the future, where agents and people work side by side. A strong software engineer who wants to own complete solutions rather than one piece of a larger system. You'll be expected to figure out how to solve the problem and make it work.
The work includes two aspects:
Vertical solutions — build and ship agents and applications for teams across Windward. Often you'll be the sole engineer on a project — scoping it, architecting it, shipping it to production.
Org-wide solutions — build the shared products and infrastructure that AI runs on across Windward. The work compounds: what you ship here makes every downstream agentic effort faster and better.
Your Mission
Build the internal agentic capabilities that change how Windward works — both the org-wide solutions that everything else connects to, and the vertical solutions that make our teams measurably better at their jobs.
How The Team Works
We move fast, ship continuously, and care more about outcomes than titles or process. If you have a bias for action, enjoy being hands-on, and naturally prioritize value, this is the team for you.
What you'll do
Take a problem from "fuzzy" to "shipped" — work from a rough definition and a priority, and figure out the rest.
Architect, build, and ship AI agents, applications, and intelligent workflows end-to-end. Integrate AI solutions with company data, APIs, and internal systems.
Rapidly prototype and experiment with new AI technologies, turning successful concepts into reliable production solutions.
Embed with internal users — sit with the team, watch how they work, map the process, build what fixes the highest-leverage part.
Own the build end-to-end — architecture, implementation, evals, logging, deployment.
Debug the non-deterministic — when a system fails on edge cases, you build the evals and guardrails to catch it rather than reading logs by hand.
Stay honest about what AI can and can't do — scope realistically, ship reliably, don't overpromise.
Requirements
First and foremost, you’re a software engineer who loves building.
3+ years as a software engineer. You've shipped products that people used.
You code AI-native — Claude Code, Cursor, Codex, whatever you've landed on — and you can read what the model writes and tell whether it's good. Python and Node are your regulars.
An all-around full-stack builder, not a coder. You have a product mindset, can hold a complex platform architecture in your head, and you can also walk into a department, talk to users, and ship a solution without a detailed ticket telling you what to build.
Hands-on experience designing, building, and evaluating production-grade LLM and agentic AI systems, including tool calling, RAG, MCP, workflow orchestration, and evaluation frameworks.
You have great intuition and pragmatism, knowing when to build, integrate, automate, or leverage existing tools. Curiosity and a strong bias toward experimentation and shipping.
Strong understanding of APIs, databases, integrations, and cloud environments. You can scope a system without hand-waving.
English at a professional level.
Nice to have
You've founded a company or built something solo that real people used.
Experience with workflow automation and integration platforms (e.g., n8n, Workato, or similar).