We are looking for an experienced Data Scientist to join our growing machine learning (ML) team at PerfectRec.


About PerfectRec

PerfectRec is a machine-learning powered decision engine focused on helping people find good, personalized recommendations without exhaustive research.  Our goals are to get each person the best product for them, while saving time and money.  We aim to attract billions of users, collectively save them billions and billions of hours on their product searches, and provide a similar amount of value by guiding our users to the right products for themselves at the lowest price.

We are starting with smartphones, laptops and other consumer electronics since there is an immediate and growing need for reliable recommendations for these products. We then plan to recommend other products such as cars and clothing, and then ultimately we’ll expand into other, more complicated decisions such as where to live or work. (Since you’re reading this, our non-ML based recommendation is that you really consider working here.)

Spammy search results, fake reviews, and paid affiliate sites make it nearly impossible to get an honest and relevant product recommendation. Finding the right item to meet your unique needs is tricky. Making the wrong choice based on bad information is a frustrating waste of time and money. We’re solving the problem by combining actual human subject-area experts with machine learning to help everyone — regardless of their expertise — find their perfect product, quickly.  To do that, we guide users through a series of questions about what they’re looking for. After each answer, we share updated recommendations and encourage answering more questions to further improve our recommendations. We get more information about a user’s specific needs from our quick and easy multiple-choice questions than e-commerce sites and search engines can deduce from search queries and past browsing data.  Answers to our questions allow us to make better recommendations tailored to each user’s unique needs.


A Small Team Where You Can Have A Big Impact

PerfectRec is a small, early-stage startup where you can have a big impact. Our founding team has experience building successful startups in the e-commerce and recommender systems space, as well as working at big organizations, like Google, Amazon, Microsoft and UC Berkeley. We are growing fast, and looking for people who are excited to build a new product where they can have a big impact. Our team is full of people passionate about the products we support, and recommending them.


Founder Led, Founder Funded

Our CEO Joe Golden bootstrapped his first startup, the custom photo products e-commerce site Collage.com, to ~$100m in revenue and a successful exit in March 2021. He’s using proceeds from that sale to build PerfectRec.


Working At PerfectRec

We are an all-remote company, so we don’t care where you live as long as you work well in a fast-paced, collaborative remote environment.

PerfectRec is an Equal Opportunity Employer. We are committed to finding the best candidates for every role and giving them the tools and guidance they need to succeed. Our all-remote structure enables us to recruit anywhere in the world, so our talent pool isn’t limited to a few big cities, and our asynchronous structure allows our team to work when it’s best for them. We celebrate diversity and are committed to creating an inclusive environment for all employees.

Learn more about how we work and our differentiating values.


What You’ll Do

On the ML team you will work on improving and expanding the core decision engine that powers the recommendations at PerfectRec. You will do this by identifying and analyzing problems as well as opportunities, exploring solutions and running experiments, and finally writing the code that puts your solution into practice and taking it live to users. Your work will directly impact and improve the user experience of anyone using the PerfectRec service and help countless users find the perfect products!

One of the areas you will work on is investigating and fixing bad predictions. When the model makes a not-so-perfect recommendation you will investigate deep into the model why this happened using analysis tools like Shapley values and others. You will identify what the likely cause is, this could be a data, feature or model issue among other things, and implement solutions that fix the problem. After running offline experiments to confirm your solution works you will roll out your improvements to production and see the recommendations improve on the live site.

  • Create and explore datasets for powering recommendations, working closely with domain experts.
  • Prototype models for making recommendations in new domains.
  • Use latest ML technologies such as Neural Networks, Transformers, XGBoost
  • Evaluate and compare different models, analyzing their tradeoffs in order to decide which model to use.
  • Add new capabilities that are shared between models in different categories.
  • Solve technical problems across the stack and contribute to all parts of our code base, including creation of our internal machine learning library.
  • Share your expertise and experience to grow with the team.
  • Collaborate with product, design, and engineers to understand customer needs, design solutions, and prototype, launch, and iterate.
  • Build a product that is genuinely loved by users and respects their time.


What You’ll Bring

  • Strong sense of ownership. Ownership is taking responsibility for the outcomes of a task or project. You’ll relish these ownership opportunities and ensure that tasks, projects, or the team are making positive forward progress.
  • Clear and concise communication. Building software is a team sport that relies heavily on great communication. You’ll be responsible for clear, thoughtful, and concise written and verbal communication. This includes the ability to give and receive feedback with empathy and understanding whether in a code review or a 1:1.
  • Willingness to learn and teach. We’re always evolving with new tools and practices. You’ll be willing to jump in and learn unfamiliar technologies while also bringing others along with you. Mentorship and helping others learn are important parts of achieving this.
  • Creative problem solving. Creating a decision engine that can make recommendations across wide ranging domains is hard, and the potential approaches are limitless. You’ll bring a research mindset and scientific rigor to evaluate and explore existing solutions, suggest new approaches, and be able to build models that successfully put them into practice.
  • Balance of perfection and pragmatism. Software has a lot of trade-offs and ways to solve problems. You’ll be able to recognize and balance the trade-offs between long-term system health and bringing value to customers faster. Your tooling choices will reflect ability to solve the problem the best way and not just the latest and greatest.



We believe that everybody has the ability to learn and grow. If you’ve used other tools, languages, and frameworks that are similar to what is listed, please apply!

  • MSc in a relevant field (computer science, ML, statistics, math, etc.)
  • 2 years of work experience
  • Experience applying statistical models and/or machine learning to real-world problems
  • Python or at least one programming language such as Java/C/C++
  • Knowledge of classical machine learning techniques such as classification, regression, regularization, and cross-validation.
  • Proficient in exploratory data analysis and data visualization.
  • Excellent team-work and communication skills
  • You’re thought about the problems we’re solving and would find them extremely interesting


  • PhD
  • 3 or more years of work experience
  • Knowledge of Recommender Systems
  • Experience writing production-quality code, using version control, making and reviewing pull requests
  • sklearn, Keras (or other deep learning frameworks like Tensorflow and PyTorch).


Salary & Benefits

We believe in offering generous compensation and benefits tailored to our all-remote workplace. We pay well because we want to work with highly motivated people that will help make the company successful.


Pay Range and Level of Role

$130,000 – $300,000 salary and stock options which vest over 4 years.  We are hiring data scientists with different levels of experience.  Compensation is based on background and experience.



  • 100% remote and no cost-of-living pay cuts.  We want you to work where you work best and we are not going to “adjust” your pay if you move somewhere with a lower cost of living.
  • We pay 100% of your and your dependents’ healthcare premium in a high-quality, national PPO plan for US-based employees. For international employees, we will work with you if you’re not covered by your national health plan.
  • You’ll get to use whatever laptop will work best for you.  You’re welcome to use your own, or a company laptop.  Of course, we recommend using our laptop recommender to pick.
  • Flexible PTO. We trust everyone to take time off appropriately for themselves. We’re pro-regular vacations – not taking regular breaks isn’t good for individuals or the team.
  • Worry-free sick days. If you’re sick, take the time off to get well. Just let your manager know as soon as practicable.

Tagged as: 1-3 Years, C#, Java, Python