> Markdown version of [/jobs/ext/3778907-software-engineer-applied-ai-hybrid](https://www.wearedevelopers.com/jobs/ext/3778907-software-engineer-applied-ai-hybrid). Every page supports `.md` or `Accept: text/markdown`. Links point to the HTML versions so they work for humans too. Agent guide: [/agents.md](https://www.wearedevelopers.com/agents.md).

---

# Software Engineer, Applied AI (Hybrid

- **Company:** TAKEONE NETWORK CORP.
- **Location:** Seattle, WA, United States (Remote available)
- **Experience:** Expert
- **Salary:** $18,000.0 - $24,300.0
- **Contract:** Permanent contract
- **Skills:** LangGraph Framework, Java (Programming Language), Computer-Aided Design, Amazon Web Services, Data Security, Python (Programming Language), Node.Js, TypeScript, LangChain, Large Language Models, Multi-Agent Systems, Llamaindex, Backend, Agentic-AI, Build Management, Amazon Bedrock, AutoGen, Claude, GPT
- **Published:** October 16, 2026
- **Apply:** https://startup.jobs/senior-software-engineer-applied-ai-wrapbook-3-10484569

## About the Role

* 6+ years of back-end development experience with Python, TypeScript/Node.js or Java, with a track record of delivering systems built for security, scalability, availability and performance
 * Hands-on experience building and shipping LLM-powered agents or workflows (with models such as Claude, GPT, Llama or Mistral), including prompting, tool use and structured outputs. You can walk us through one you built: how it was designed, what broke and how you fixed it
 * Production experience with at least one agent or LLM framework, such as LangGraph, LangChain, LlamaIndex or AutoGen, and a solid understanding of AI architectures like RAG, reranking and multi-agent orchestration. Experience with LangGraph specifically is a plus
 * A security-first mindset: you think about data access, permissions and failure modes in everything you build, especially when agents can take real actions
 * Strong product sense and communication skills, with the ability to work directly with non-technical teams to understand their workflows and turn them into reliable automation
 * Comfort with ambiguity and an ownership mentality: you can take a loosely defined problem from idea to production on your own
 * Nice to have: experience running AWS workloads in production and knowing how to design for, deploy on and get the most out of Amazon's services, ideally including Amazon Bedrock AgentCore

## Description

* Design and build production-grade AI agents with LangGraph that automate multi-step workflows for internal teams across Wrapbook
 * Deploy, run and scale agents on Amazon Bedrock AgentCore, using its runtime, memory, identity and gateway capabilities to make them secure and reliable, while tracking what they cost to run and keeping it efficient as usage grows
 * Partner with internal teams to find high-impact workflows, draw out requirements from people who can't hand you a spec, and turn them into agent-driven solutions, explaining clearly what each agent can and can't reliably do
 * Integrate agents with Wrapbook's internal systems, APIs and third-party tools so they can take real actions, not just answer questions
 * Build evaluation, observability and guardrail frameworks to measure agent quality, catch failures and keep humans in the loop where it matters
 * Establish reusable patterns, shared tools and best practices that let the Agentic Systems team ship new agents quickly and consistently
 * Own agents end to end, from prototype to production, measuring their impact on the teams they serve and iterating based on usage data and feedback
 * Drive adoption through clear documentation, training and enablement, so teams understand what each agent does, trust it and know how to get the most from it