Data & AI Engineer Intern

Midtronics•Willowbrook, CA

1d•Hybrid

About The Position

Midtronics is the global leader in battery management technology, serving automotive OEMs, dealerships, fleets, and energy storage markets worldwide. Our products and platforms help customers make better decisions about batteries, from manufacturing through end-of-life. The Data Analytics & AI team is a small, high-impact group that owns the company's data infrastructure, analytics platforms, and AI initiatives. We build the dashboards executives use to make decisions, the ML models embedded in production hardware, and the internal tools that make every department faster. We operate with a startup mindset inside a mid-size company, which means you'll ship real things, not just prototypes. We're looking for a Data & AI Engineer who can move fluidly between building AI-powered tools, wrangling messy enterprise data, designing compelling dashboards, and sitting in a room with business stakeholders to figure out what actually matters. This is not a pure data engineering role or a pure analytics role. It's a builder role for someone who is equal parts technical and creative, and who gets energy from turning ambiguous business problems into working solutions.

Requirements

3+ years of experience in data analytics, data engineering, or a similar technical role
Proficiency in Python or Node.js for scripting, automation, and building tools
Hands-on experience with Power BI (DAX, data modeling, report design) or equivalent BI tools
Strong SQL skills and experience working with relational databases or structured data sources
Demonstrated ability to work with APIs and integrate data across multiple systems
Clear, effective communication skills with both technical and non-technical audiences
A portfolio, GitHub, or track record of things you've built that you're proud of

Nice To Haves

Experience building applications or automations using LLM APIs (Claude, OpenAI, etc.)
Familiarity with cloud platforms (Azure preferred), CI/CD, and version control (Git)
Experience with Smartsheet, Epicor, or other enterprise SaaS/ERP platforms
Background in manufacturing, automotive, or B2B enterprise environments
An eye for design and UX, whether that's a dashboard layout, a slide deck, or a data visualization
Comfort with ambiguity and a bias toward action over perfection

Responsibilities

Design, develop, and deploy AI-powered tools and agents using LLM APIs (Anthropic Claude, OpenAI) with structured tool-use patterns
Build retrieval-augmented generation (RAG) pipelines, prompt engineering workflows, and evaluation frameworks
Integrate AI capabilities into existing business processes such as sales forecasting, trip reporting, and document generation
Stay current on AI tooling and bring new ideas to the team about where AI can create real leverage
Build and maintain data pipelines connecting enterprise systems (Smartsheet, Epicor ERP, SharePoint, JIRA, GitHub) into unified, queryable data sets
Perform exploratory and diagnostic analysis on sales, operations, and product data to surface actionable insights
Own data quality: identify gaps, reconcile conflicting sources, and build automated validation checks
Work with SQL, Python (pandas, NumPy), and APIs to extract, transform, and load data at scale
Build and maintain interactive Power BI dashboards for sales performance, pipeline health, and operational KPIs
Design data visualizations that tell a clear story and drive decisions, not just display numbers
Create self-service reporting tools that empower stakeholders to explore data without hand-holding
Develop HTML-based interactive dashboards and artifacts when Power BI isn't the right fit
Partner directly with Sales, Marketing, Engineering, and leadership to understand their real problems (not just their data requests)
Translate vague asks like 'we need better visibility into the pipeline' into concrete, scoped deliverables
Present findings and recommendations to non-technical stakeholders in a way that's clear, concise, and actionable
Bring a creative, 80/20 mindset: find the fastest path to a useful answer before overengineering