About this role
Full-Stack Developer – AI Systems & Agent Engineering
Role Summary
Direct Agents is running a growing fleet of AI-powered systems across our client and internal reporting stack — automated account agents, intelligence systems, always-on client dashboards — and is now building a new, confidential generative AI product for the media industry.
We're hiring a full-time, remote Full-Stack Developer to do both: help keep the existing fleet running and improving, and help build the new product from the ground up. A meaningful part of this job is working inside Claude Code as our primary engineering environment — steering and reviewing AI-generated code and agent workflows, not only writing everything by hand.
What You'll Work On Live systems you'll help support and extend:
- Internal agent platform — syncs and synthesizes account data, call transcripts, and performance signals across every client account, surfacing findings automatically
- Client-facing report sites and dashboards — hosted, JavaScript front ends backed by a SQL data layer and marketing-data pipelines, live for a dozen-plus client and internal audiences
- Workflow automations — connecting data sources, chat, and email/calendar tools so reporting and account data move without manual steps Plus:
- The confidential new product — a from-scratch generative AI build for the media industry, where you'll help shape architecture and features from day one
Key Responsibilities
-
Build and maintain full-stack features across our Python/FastAPI backends, JavaScript front ends (Vue preferred), and SQL data layer
-
Work inside Claude Code daily: prompt and steer agentic coding workflows, review AI-generated code and pull requests for correctness, and extend our library of custom skills and subagents
-
Debug and harden the data pipelines (marketing-data extracts, SQL views, workflow automations) that feed client dashboards and internal agents
-
Integrate LLM APIs (Claude, GPT) for generative features — copy generation, insight extraction, creative/image analysis — in both the existing systems and the new product
-
Own deploys for live client-facing sites and internal tools, including monitoring, rollback, and incident response
-
Collaborate with data, strategy, and account teams to translate reporting and media needs into working software
-
Document systems, prompts, and workflows so institutional knowledge isn't locked in one person's head
-
Contribute to architecture and feature decisions on the new confidential product alongside day-to-day support work
-
Our Stack & Tools
-
Primary dev environment: Claude Code — agentic coding, subagent workflows, custom skills. Comfort working AI-assisted (directing and reviewing agent output, not only typing every line by hand) is a core requirement for this role, not a bonus skill.
-
Backend: Python, FastAPI, SQL (SQL Server / Azure SQL)
-
Frontend: Vue.js (React or Angular experience also useful — several of our tools are React)
-
Data & pipelines: marketing-data platforms, workflow automation tools (e.g. n8n), SQL databases
-
Deployment & infra: Netlify, containerized cloud services, Git-based version control
-
AI: Claude and GPT APIs, retrieval/vector search, agent orchestration patterns
Qualifications
- Bachelor's degree in Computer Science, Engineering, or a related field (or equivalent experience)
- Proven full-stack development experience, with genuine interest or hands-on experience in AI-assisted or agentic development — not just calling an LLM API, but working inside tools like Claude Code, Cursor, or similar
- Solid backend skills in Python; front-end proficiency in JavaScript, Vue preferred
- Familiarity with LLM APIs, vector search, or media/marketing AI workflows is a strong plus
- Experience with RESTful APIs, Git-based version control, and agile workflows
- Comfortable owning both greenfield build work and ongoing support/maintenance of live, client-facing systems
- Excellent collaboration and communication skills — this role works directly with data, strategy, and account teams, not only engineering