Founders get an enormous amount of support at the start — accelerators, investors, communities, playbooks for every stage. At the end, they get almost nothing.
Starcycle exists for that part of the journey. We help founders shut down their companies faster and with dignity, turning months of legal and administrative grind into several painless days.
Every one of those companies also comes with records of real people, doing real work, under real constraints. Protecting the people inside those records is where this role begins.
We’ve spent the past two years building trust, brick by brick, by sitting with founders and teams on some of the worst days of their lives. Companies trust us with this at their lowest moments. To be able to do this work is an honor.
As Starcycle’s Data Analyst, you’ll be the human ground truth for our PII and data-quality work. You’ll play a fundamental role in shaping how our high bar translates into scalable judgment, helping build and scale pattern detection, yet you’re not above reading a thousand rows by hand if that’s what the task calls for.
🤓 About you
We expect you to be an entrepreneur in your own right: passion for using technology to solve difficult problems; willingness to roll up your sleeves and make an impact on day one; and saying yes and taking ownership.
- You’re biased towards action; you’re always ready to take initiative. “This is not my job” is not part of your vocabulary.
- You cope well with frequent change and ambiguity, and can shift gears quickly and decisively, with a default bias towards urgency without losing precision.
- You’re a team player. Working together as a team, sharing knowledge, and communicating with empathy and clarity comes naturally to you. Your instinct is to help and contribute, not just sit on the sidelines.
- You’re warm and empathetic. You’re the one in your friend circles that everybody trusts, and you know exactly what that trust costs.
- You’re naturally curious, excited about solving problems, and love learning on the fly. A new problem or question presents itself to you as an opportunity to learn and grow, rather than a roadblock.
- You communicate proactively: status updates before anyone asks, problems the moment you see them (with a proposed next step), questions before they snowball, and edge cases or other unusual flags before they spiral.
- You have real intuition for machine learning, beyond the classroom or the research papers, and can confidently stand behind all of your work and explain it line by line regardless of audience.
You’ll be rolling up your sleeves and diving in alongside the entire team, working and shipping against the high bar that our clients have learned to expect from us.
📝 Your responsibilities
These are indicative, not exhaustive, and may change over time as our client and project needs evolve.
- Contribute to our systems that label PII across raw company text at scale, and help define what correct means
- Adjudicate the ambiguous cases, the ones reasonable people disagree on, and turn those calls into the standard reference, building the held-out sets every detector is scored on
- Measure detection by tracking false flags and misses
- Write the detection code, regex through classifiers
- Help set the quality bar together with the founders, and own that standard as the team grows
👋 Who you are
- Fluent grasp of ML and NLP fundamentals with strong intuition
- Ability to communicate clearly and concisely in English, with native- or bilingual-level understanding of conversational context and nuances
- Fluent with AI tools, with deep understanding of prompting and building without solely relying on AI
- Comfort with ambiguity and short feedback loops
Stack: Python and SQL, preferably used on real data and not just in the classroom
😶 You’re probably not a fit if…
- You want fully defined projects, tickets, and queues
- You’d rather build the model than hold the model accountable
- Reading data closely for hours sounds like a punishment rather than a puzzle
- You’re only interested in strategy and not the hard, hands-on work in the trenches
🎁 Extra points if…
- You have hands-on data labeling, annotation, or data-quality experience
- You have text-processing or information-extraction experience: regex, tokenization, entity recognition
- You’re familiar with model evaluation: precision and recall tradeoffs
- You have prior exposure to privacy-, security-, or compliance-sensitive data handling
- You have prior experience at an early-stage startup
👥 The team
Starcycle is built by a serial founding team who has lived the problem, and the solutions, first-hand. Our CEO Jaclyn is a 3x founder with 15+ years across marketing, sales, GTM, and operations, who started Starcycle after shutting down two startups. Our CTO Red is a 2x founder and machine learning engineer who built petabyte-scale data products used by some of the nation’s largest retailers.
The rest of our operating and dev team brings experience from places like Google, Uber, Peloton, Workday, and Cash App. We’re a small, tight-knit team, proudly punching above our weight. We run through walls for the founders we work with, and carry this mentality through everything we touch.
Starcycle came about because we looked at the status quo and decided founders deserved better. And we believe the change we want to see comes from underestimated, overlooked places, by unconventional people motivated by a singular possibility: that a better future only happens if we build it.
🚀 Details
This role is a full-time contract position, currently scoped for three months. Depending on project volume and mutual fit, there is potential to extend or convert into a full-time, salaried position.
Expected duration: through December 2026
Rate: $35-60/hour, depending on experience
Location: New York City
We expect at least 50% of your time with us to be in-person and are currently not able to support fully remote candidates.
📨 How to apply
Send an email to founders@starcycle.ai on the worst mistake you’ve ever made, how you salvaged it (or not), and what you learned.
No traditional cover letters please; this email is an open canvas. Treat it as your first point of evaluation.