Senior Full-Stack Engineer, AI Agents

About DoWorks

DoWorks is an AWS-native SaaS platform that helps teams capture business processes, review them, and publish them for their organization. You'll work across the whole stack — frontend, serverless backend, and the AWS infrastructure that runs it.

This is a hands-on, high-ownership role on a small team where building and orchestrating AI agents is at the center of what we do. We're looking for an engineer who designs agentic systems, not just someone who prompts a chatbot.

What makes this role different?

Agent building is the core skill for this position.

• You design and build AI agents — defining their tools, context, guardrails, and control flow — and ship them to production.
• You orchestrate multiple agents working in parallel, decomposing large goals into agent-sized tasks, delegating, and integrating the results into a coherent whole.
• You own the quality of everything agents produce: rigorous review, verification, and testing, regardless of who or what wrote the code.
• You use AI across the full SDLC — design, coding, testing, and review — and you push the ceiling on how much can be built this way.

What you'll work on

• AI agents and the tooling, orchestration, and guardrails that make them reliable.
• A web frontend and a serverless TypeScript backend, sharing a typed contract layer.
• A GraphQL API layer over serverless compute.
• A data layer spanning relational (PostgreSQL) and NoSQL (DynamoDB) stores, with tenant isolation.
• The AWS infrastructure (CDK) that deploys it all.

Tech stack

• Agent runtime (required): the full Amazon Bedrock AgentCore stack — AgentCore Runtime, AgentCore Gateway, and MCP tool integration.
• Frontend: React, SvelteKit, TypeScript
• API: GraphQL, AWS Lambda + API Gateway, TypeScript
• Data: Aurora PostgreSQL, DynamoDB, Amazon Cognito
• Infra: AWS CDK, TypeScript

What we're looking for

• Proven experience building AI agents — tool design, context management, guardrails, and control flow — and shipping them to production.
• Hands-on production experience with the full Amazon Bedrock AgentCore stack — Runtime, Gateway, Identity and MCP-based tool integration.
• Hands-on experience orchestrating multiple AI coding agents in parallel across real workstreams.
• Strong TypeScript and solid backend fundamentals.
• Real AWS experience (Lambda, databases, auth) and infrastructure-as-code.
• GraphQL API design experience.
• Comfort modeling data across relational and NoSQL stores.
• Comfort with multi-tenant auth, security, and secrets handled correctly.
• A testing-first mindset — every change ships with tests.