Senior Data Scientist – Public Sector Finance Analytics

Senior Data Scientist

Element
14 is hiring a Senior Data Scientist to do the analytical core of our financial analytics work for a public sector organization focused on housing and community development. You will own real questions end-to-end — from raw transactional data through to models and analytical products that program staff and leadership use to make better decisions. This is not an academic seat.

We are looking for someone who has shipped real work on hard problems, can defend their methods, and would rather solve a problem that matters than write the perfect paper about it.

Element
14 is building a production financial analytics capability for a public sector organization focused on housing and community development. The organization manages a large grant and program portfolio, and the team's job is to help leadership and program offices understand it — where the money is going, how programs are performing, where the patterns in the data warrant attention, and what the modeling work can do to support better decisions.

The team produces three flavors of analytics, each with its own users and its own data. Portfolio analytics gives leadership the macro view of the program landscape. Entity analytics gives program officers and analysts the per-firm view that supports program management decisions. Transaction analytics gives operating staff a per-payment view at the point of action. We start producing real value on day one, using the organization's own data and public sources.

The work is data science and engineering applied to financial data. Building statistical and ML models on disbursement and recipient data. Designing analytical products that program staff and leadership actually use. Connecting the organization's internal systems to public datasets in ways that reveal patterns the organization could not see on its own. Careful, defensible work that holds up under scrutiny.

We are AI-first by default. Modern data science and ML are part of the toolkit on every engagement — LLMs for parsing unstructured documents and free-text fields, ML for scoring and classification, agentic workflows for repetitive analytical work. We are not chasing AI for its own sake, but we are not doing 2018-era data science either. We expect the people on this team to be fluent in current tools and to use them to be faster and sharper than the consulting median.

The data sources span the organization's own systems, commercial data, and public records. Organizational systems include the general ledger, grant tracking, and program disbursement systems. Commercial sources include major entity and identity data providers. Public sources include USA Spending.gov, FFATA sub-awards, SAM.gov, and IRS Form 990s. The interesting analytical questions almost always live at the intersection.

Beyond this engagement, we expect this team to grow with the firm. As Element
14 wins additional federal and state work, the people we hire now will help shape future engagements and the capabilities we build.

• Build statistical and ML models on the organization's financial and program data. Models that describe what is happening in the portfolio, models that predict where attention is most needed, models that explain variation across programs and recipients. Make the factors interpretable to a program officer who needs to act on the output.
• Apply classical statistical methods alongside modern ML. Time series analysis on disbursement patterns, regression and causal inference where they fit, supervised and unsupervised methods where they earn their keep. Choose the right tool for the question, not the trendiest one.
• Work with the full data landscape — internal transactional data, commercial entity and identity data, public records. Cross-reference, link, and reason across sources. Entity resolution is part of the craft when working with messy real-world data, where the same recipient often appears in different forms across datasets.
• Use modern AI tooling as a standard part of the workflow. LLMs to extract structure from PDFs, narratives, and free-text fields; agentic workflows for repetitive analytical work;…