Quantitative research intern

About Our Company:

We're a software company in the US working on innovative projects that help businesses and individuals thrive. Our team is passionate about exploring new ideas, solving complex problems, and creating technology that makes a real difference. We value learning, growth, and collaboration, and we're committed to making great things happen with our tech.

The Role:

We're looking for Quantitative Research Interns to join a six-month research cohort. You'll take on a focused research question and work through it with real data, inside a setup built for running a lot of small experiments quickly.

The loop is the same every time: form a hypothesis, test it cheaply, then try hard to prove yourself wrong. Most ideas don't survive that, and that's the expected outcome, not a bad week. We measure how many questions you resolve (including the ones you correctly rule out), not how many wins you find. If a cleanly discarded idea feels like wasted effort to you, this role won't suit you. If it feels like progress, read on.

This is a research apprenticeship. You'll get a bounded question, real data, and access to our proprietary research platform, the same environment the full-time team uses.

What You'll Do:

Turn observations into testable hypotheses with a stated mechanism, a predicted direction, and the conditions that would prove you wrong, written down before you run anything. Run cheap first-pass tests before expensive ones, and retire weak ideas quickly. Hunt for the ways you might be fooling yourself: leakage, survivorship, samples too small to mean anything, results driven by one slice of the data, multiple testing. Build and run your analysis on our proprietary research platform, using built-in AI assistance for tutoring, code, confounder brainstorming, and adversarial critique. AI use is expected and coached here, not policed. Defend surviving work in review, then write it up as one page somebody else can reproduce. Log every experiment, failures included, so the next researcher doesn't repeat your dead end.

Week one is a guided tutorial experiment, so you learn the loop before you take on a real question.

What We Need From You:

Be Honest: Always search for the truth and report the result you actually got, especially when it kills your favorite idea. Work Well With Others: Be open to learning from everyone. Your failed experiments are shared team assets, so write them up for the person who comes after you. Be Open to Challenges: Expect your work to be challenged in review. The right response is to update, not to defend. Make Smart Decisions: Know which idea deserves another day and which one is finished. Good judgment about when to stop is worth more than raw speed. Understand Challenges: Recognize and address the problems in front of you, whether they're deadlines, messy data, or a result that looks too good to be true. Be Persistent: A month of ideas that don't pan out is normal. We're looking for people whose curiosity survives that.

Skills & Qualifications:

Working toward, or holding, a degree in a quantitative field (statistics, economics, mathematics, physics, computer science, or engineering), or a self-taught equivalent you can show us. Comfortable with Python and pandas, and willing to work on our proprietary research platform. Able to read and debug somebody else's code. SQL is a plus. A solid grasp of basic statistics and the ways it goes wrong: sample size, multiple comparisons, overfitting, correlation without a mechanism. Evidence of self-directed learning: a project, a repo, a write-up, or a question you chased down because it bothered you. Thoughtful, fluent use of AI tools, and the judgment to check what they hand you. Clear written communication. A large part of this job is explaining a result honestly in one page. Prior industry or domain experience is welcome but not required. We'll teach you what you need. Curiosity, reasoning, and the instinct to check your own work count for far more.

What You'll Walk Away With:

A documented, reproducible record of six months of real research, including the work you ruled out and why. Working fluency in experimental design, data hygiene, and reproducibility on real data, the parts that are hard to learn outside a working team. Serious practice at AI-leveraged research, which is quickly becoming the core skill in this field. A reference from people who watched you work closely for six months, and a first look whenever we open a full-time seat.

Why Join Us:

Be part of a team that values honesty, teamwork, and innovative technology. Work on meaningful projects that have a real impact. A supportive environment that encourages your professional growth. A flexible, remote-friendly schedule. Real data, our proprietary research platform, and direct review of your work by the people who built it.

Program Details:

Length: 6 months, cohort-based. Hours: Full-time preferred; part-time (minimum 20 hours/week) considered for students with a fixed schedule. Location: Remote, with meaningful overlap with US working hours. Compensation: Paid. Start date: As soon as possible. Strong performers may be considered for extension or a full-time role.

We're committed to creating a diverse and inclusive workplace. Everyone is welcome to apply, regardless of your background.

We're excited to see what you can bring to our team and how we can make great things happen together.