Software Developer Intern — AI Tools Team

Winchoice; Arizona; On-site; USD 16,000 - 19,000; Full time

Job summary

WinChoice USA offers a hands-on internship where you’ll ship production tools used by our sales, fulfillment, and executives. You’ll contribute to internal tooling, AI agent infrastructure, and live-data integrations, with production deployments and real-time feedback.

Expect to work 20 hours per week at $15/hour on-site in Arizona, Spring 2027, with potential extension into Summer. You’ll collaborate across teams and learn how real systems operate, not just demos.

Qualifications

• Proficient in Python, JavaScript/TypeScript, or similar language.
• Familiar with APIs: reading docs, auth, and handling unexpected responses.
• Git version control is part of your daily workflow.
• Able to explain your code to non-developers.

Responsibilities

• Build internal tools that automate real operational tasks.
• Collaborate with AI team on agent-facing infrastructure.
• Integrate with live systems like HubSpot, BigQuery, Graph.
• Write automations and data pipelines that run unattended.
• Engage in requirements discussions to capture the business problem.

Skills

Python

JavaScript/TypeScript

APIs

Git

Code explanation

Tools

HubSpot

BigQuery

Microsoft Graph

Job description

Description

You'll ship things people actually use. By people, we mean the 400 humans down the hall.

Most developer internships hand you a sandbox project nobody will ever run. You build something, you demo it in August, it gets deleted in September.

That's not this.

WinChoice USA is a 400+ person home improvement company, and our IT team builds its own software. Not scripts; actual internal tools and AI systems that the sales floor, the fulfillment team, and the executives depend on every day. We've built MCP servers that let AI agents query our real business systems. We've built automation that moves data between HubSpot, BigQuery, and a half-dozen SaaS platforms. We've built reporting that the CEO looks at on Monday morning.

You'll work on that. The thing you build will be in production, and someone will complain when it breaks. That's the job.

What you'll be doing

• Building internal tools that solve real operational problem, the kind where someone is currently doing something manually and hates it.
• Working alongside our AI team on agent-facing infrastructure: MCP servers, tool definitions, and the plumbing that lets an LLM do useful work against real data.
• Integrating with live systems: HubSpot CRM, BigQuery, Microsoft Graph, and whatever else the problem requires.
• Writing automations and data pipelines that run unattended.
• Sitting in the requirements conversations, not just receiving a ticket. You'll hear the business problem before it becomes a spec.

Requirements

What we need from you

• You can write code in Python, JavaScript/TypeScript, or something close enough that you'll pick it up fast.
• You understand APIs — how to read docs, authenticate, handle a response that isn't what you expected.
• Git doesn't scare you.
• You can explain what your code does to someone who doesn't write code. This matters more here than you'd think.

We're not expecting a CS degree finished or a portfolio of polished apps. What'll make you stand out

• You've built something with an LLM API and have opinions about what worked and what didn't.
• SQL. Real SQL, not just SELECT *.
• Cloud exposure: AWS, Azure, containers, anything.
• Something you built because you were annoyed by a problem. Those are our favorite projects too.

Why this one is worth your spring

Two reasons.

First, the AI work here is not a demo. A lot of companies are doing AI theater right now, a chatbot on the website, a pilot that never ships. We're building agent infrastructure against live business data because it saves real hours. You'll learn what actually works at that layer, which is knowledge most people your age can't get yet because most places aren't doing it.

Second, you'll see the whole line. Business problem ? conversation with the person who has it ? design ? build ? deploy ? the phone call when it breaks. At a big company you'd own one slice of that for two years. Here you'll see all of it in a few months, and you'll find out whether you like building products or building systems.

You'll be working directly with a team that built this department from nothing, including the person running it. Ask hard questions. We'll answer them.

Details Spring 2027 with possible extension into Summer

Up to 20 hours per week $15/hour 100% Onsite