Senior Data Scientist, Casual Inference & Experimentation in United

Job Description

About the company

Harnham is partnering with a seed stage, founder led technology company building a decision intelligence and optimization platform for consumer brands.

Many CPG and retail organizations have more data than ever, yet much of it is still used primarily to explain what already happened. The company is building technology designed to help teams understand what is actually driving business outcomes, measure the impact of commercial decisions, and determine where to act next.

The platform brings together causal inference, experimentation, machine learning, and applied statistical modeling to turn complex commercial data into recommendations that customers can use to make better decisions.

The founders spent significant time working directly with brands and validating the problem before formally building the company. Today, the business is seed funded with a team of eight, including two Data Scientists, and is entering its next phase of product development and customer adoption.

An intentionally early stage environment gives the person joining an opportunity to help shape the product, modeling infrastructure, customer use cases, and internal processes as the company grows.

The role

Harnham is seeking an experienced Data Scientist to take ownership of a core area of the measurement and modeling work behind the product.

The position sits at the intersection of causal inference, experimentation, production Data Science, and customer problem solving. You will take methodologies from research and prototyping into production, validate them against real customer data, and help determine how model outputs translate into recommendations customers can act on.

You will have significant ownership from the beginning, along with direct access to the founders, customers, Data Science team, engineers, and academic advisors.

What you will do

• Design, test, validate, and improve Python based statistical and machine learning models
• Apply Synthetic Control, Difference in , and Double Machine Learning to real world commercial problems
• Take causal inference and experimentation work from research and prototyping into production
• Design and evaluate A/B tests and experimentation frameworks
• Validate model performance using customer data and identify areas where methodologies can improve
• Own a core modeling workstream and be accountable for shipping it into production
• Support customer POCs and work directly with customer stakeholders
• Translate complex statistical findings into clear recommendations for technical and nontechnical audiences
• Collaborate closely with the founders, Data Scientists, engineers, advisors, and customers
• Help establish technical standards and ways of working as the company grows

What Harnham is looking for

A PhD is strongly , with strong Master’s candidates also considered, in Economics, Statistics, Computer Science, or another highly quantitative discipline.

Candidates should also bring:

• 1 to 2+ years of professional experience applying causal inference and experimentation methodologies in production
• Hands on experience with Synthetic Control, Difference in , and Double Machine Learning
• Strong A/B testing and experimentation experience
• Strong Python experience designing, testing, validating, and improving statistical or machine learning models
• Experience taking analytical or machine learning work beyond research and into a real product, production environment, or customer use case
• The ability to communicate sophisticated technical concepts clearly to both technical and nontechnical audiences
• Comfort working directly with customers and senior stakeholders
• A high level of ownership and comfort working through ambiguity
• Genuine interest in building at the earliest stages of a company

Experience within technology, marketplaces, consumer, retail, or other data intensive environments is valuable but not required.

Experience building data pipelines or ingestion workflows is also a plus.

Longer term, meaningful growth potential exists for someone interested in the intersection of Data Science and product leadership. The right person could expand well beyond the initial scope of the role as the company scales.

What the environment is like

A true seed stage startup environment means ownership extends well beyond a narrow job description.

No large organization or established playbook surrounds the position. Priorities will evolve, problems will not always arrive fully defined, and each person on the team is expected to take meaningful ownership of outcomes.

The person joining will work closely with the founders, move quickly from idea to implementation, interact directly with customers, and have meaningful influence over how the product develops.

Someone who enjoys building from the ground up, working through ambiguity, and taking ownership will be well suited to the environment. Candidates looking for a narrowly defined role within a highly structured Data Science organization are unlikely to enjoy the way the team operates.

The distributed team regularly collaborates with colleagues, advisors, and customer stakeholders across U.S. time zones, along with some international team members.

Compensation and equity

The anticipated base salary range is $120,000 to $160,000, depending on experience and overall alignment with the role.

The package also includes meaningful early stage equity with a standard four year vesting schedule.

Location

The position is fully remote within the United States.

Candidates currently residing in California, Washington, or New York cannot be considered for the role.

Work authorization

Applicants must already have permanent U.S. work authorization without current or future employer sponsorship.

The company is unable to sponsor, transfer, extend, or otherwise support an employment based immigration status. Candidates relying on H-1B or H-1B transfer, F-1 OPT/STEM OPT or CPT, TN, O-1, E-3, L-1, J-1, H-4 EAD, or other temporary visa or EAD based work authorization cannot be considered for the search.