Vibe Coder (AI Developer)

Harmon Digital · Austin TX

We’re looking for a Vibe Coder—an AI-savvy builder who blends sharp product instinct with fast, thoughtful engineering. You ship early, iterate often, and make tools feel alive. You care about the experience as much as the architecture.

What you’ll do

• Design, build, and ship AI-powered workflows: agents, automations, and micro‑apps that actually move business metrics.
• Craft prompts, retrieval pipelines, and guardrails; evaluate outputs and reduce hallucinations with data‑backed methods.
• Integrate third‑party APIs and internal systems; stitch together clean data flows across CRMs, ATS, databases, and queues.
• Stand up durable services: logging, observability, error handling, and performance tuning for production reliability.
• Collaborate with product and ops to translate messy, real‑world processes into elegant, maintainable systems.
• Document decisions, write crisp runbooks, and leave things better than you found them.

Tech you’ll likely touch

• Python or TypeScript/Node; serverless or lightweight containers; Postgres; Redis; queues.
• LLM tooling: OpenAI/Anthropic APIs, embeddings, vector stores, LangChain/LlamaIndex (or your own patterns).
• Workflow orchestration: n8n; event-driven jobs; webhooks; API integrations.
• Data plumbing: Airtable, CSV/ETL pipelines, schema design, versioned configs, secrets management.

What makes you a great fit

• You’ve shipped AI features or agentic workflows to real users (not just demos).
• Strong grasp of LLM fundamentals: context windows, function/tool calling, RAG, embeddings, prompt design, and evaluation.
• Pragmatic engineering: you balance speed with reliability; you log, monitor, and test the paths that break.
• Product sense: you notice friction, reduce it, and obsess over clarity in UX and system behavior.
• Security mindset: you handle PII with care and build with least privilege and traceability.

Nice to have

• Experience with ATS/recruiting systems or business automation.
• n8n + Airtable fluency; API-first SaaS integrations; schema consolidation.
• Lightweight MLOps: dataset hygiene, versioning, offline/online evals, simple fine-tunes when warranted.

Working details

• Location: Austin, TX (hybrid) with remote flexibility.
• Compensation: Competitive salary + benefits, aligned to experience and impact.
• Tools: Modern cloud stack, fast CI, access to leading AI models and evaluation frameworks.

How we hire

1. Intro chat — mutual fit, your work style and favorite projects.
2. Portfolio deep dive — walk us through an AI system you built, decisions and tradeoffs.
3. Build session — scoped exercise focused on prompts, retrieval, and reliability.
4. Team panel — collaboration, product thinking, and architecture discussion.
5. Offer — clear scope, growth path, and the first problems you’ll own.

If you want to craft AI systems that feel intuitive, resilient, and fast—and you love turning complex operations into simple flows—you’ll enjoy this role.

Responsibilities

• Ship AI features end‑to‑end: scope, design, implement, test, and deploy agentic workflows and micro‑apps that improve key business metrics.
• Build robust LLM systems: prompts, tool/function calling, RAG pipelines, embeddings, retrieval, guardrails, and output evaluation.
• Integrate data and services: connect CRMs/ATS/databases, design clean schemas, manage ETL, and ensure secure auth and permissions.
• Operate production reliably: logging, monitoring, tracing, performance tuning, rate‑limit and failure handling, incident response, and postmortems.
• Establish quality practices: define evaluation criteria, offline/online tests, regression checks for prompts and retrieval changes.
• Collaborate across teams: translate messy processes into elegant systems; write crisp docs/runbooks; communicate decisions and trade‑offs.
• Maintain infrastructure: CI/CD, versioned configs/secrets, lightweight containers/serverless, and cost‑aware model selection.
• Continuously iterate: instrument usage, analyze feedback, reduce friction, and simplify flows while preserving resilience.

Requirements

• Experience: 2–5 years building production software; at least 1 year shipping AI features or agent workflows to real users.
• Technical depth: Python or TypeScript/Node; Postgres; Redis; queues; REST/GraphQL; container/serverless deployment.
• LLM fundamentals: prompt design, tool/function calling, embeddings, RAG, retrieval pipelines, evaluation, hallucination reduction.
• Data plumbing: schemas, ETL/CSV handling, Airtable or similar; secrets/config versioning; clean logging and observability.
• Integrations: third‑party APIs (CRMs, ATS, billing), webhooks, event‑driven jobs; secure auth and permissioning.
• Reliability mindset: testing critical paths, monitoring, tracing, performance tuning, graceful failure handling.
• Product sense: translate messy processes into clear systems; write crisp docs/runbooks; bias for simplicity and speed.
• Security & compliance: handle PII safely; least‑privilege access; auditability.
• Communication: collaborate with ops and product; explain tradeoffs; iterate quickly from feedback.

Nice to have: n8n, Airtable, LangChain/LlamaIndex, lightweight MLOps (dataset hygiene, evals, small fine‑tunes), ATS/recruiting domain exposure.