AI / Data scientist (Performance / Model Maintenance) Location: Remote Job: AI / Data scientist (Performance / Model Maintenance) Schedule: Full-time Shift: Day Job JOB DESCRIPTION Description AcisTek is seeking qualified professionals to support the Grants.gov Information Technology Services program for the U.S. Department of Health and Human Services, Office of Grants, Division of Information and Solutions. Grants.gov is a highly visible, public-facing federal shared service that provides a centralized platform for applicants to find and apply for federal discretionary grant opportunities and for federal agencies to post, manage, and process grant application packages. The selected candidate will support the operation, maintenance, enhancement, modernization, security, performance, and user experience of the Grants.gov system. This work supports a mission-critical federal platform used by applicants, grantors, federal agencies, and partner systems across the grants lifecycle. Candidates should be comfortable working in a structured federal IT environment with strong emphasis on service continuity, stakeholder responsiveness, documentation, quality, security, accessibility, and disciplined delivery. Ideal candidates will bring experience supporting federal IT programs, public-facing web applications, grants management systems, cloud-hosted environments, Agile/SDLC delivery, system operations, application support, testing, cybersecurity, communications, or related technical and program functions. Prior experience with Grants.gov, federal grants systems, HHS, government shared services, or high-volume public-sector platforms is strongly preferred. Preferred Grants.gov / Federal Grants Experience Preferred experience includes one or more of the following: Grants.gov, GrantSolutions, or other federal grants management platforms. Federal discretionary grants lifecycle, including funding opportunity posting, application package creation, applicant submission, agency download, forms, validations, and acknowledgements. Public-facing government websites or shared service platforms with large user communities. System-to-system interfaces supporting applicant systems, grantor systems, partner systems, or external federal integrations. Federal forms development, Adobe/XDP/PDF forms, XML schemas, web forms, form validation, and Section 508 accessibility. HHS, OMB, GAO, OIG, congressional data calls, federal stakeholder reporting, or executive-level federal communications. Federal cloud, ATO, POA&M remediation, vulnerability management, disaster recovery, and operational continuity. Responsibilities include but are not limited to: Continuously track model accuracy, precision, recall, F1-score, and other KPIs to detect degradation or drift Curate and maintain gold-standard datasets, audit training data for bias or fairness gaps, and implement checks to catch data drift Schedule and execute retraining pipelines, update models with new data, and validate performance before deployment Tune hyperparameters, refine feature engineering, and apply MLOps practices to improve efficiency and scalability Deploy models into production environments (cloud platforms like AWS, GCP, Azure) and integrate with APIs, dashboards, or downstream systems Investigate performance drops or failures, identify data or model issues, and implement corrective actions Work closely with ML Engineering, Data Engineering, Product, and Legal teams to align model quality with business and compliance requirements Maintain clear records of model versions, performance trends, and maintenance activities for audit and stakeholder review. Education and Experience: Conversational analytics, missed‑intent mining, content quality scoring, generative response tuning governance, safe‑use controls Must be able to perform monthly tuning cycles Quarterly FAQ update reports Performance trend analysis impacting program budgets Degree in information technology, computer science, or related. 5 years or more of related experience required. Physical Requirements: Prolonged periods sitting at a desk and working on a computer. Must be able to lift up to 15 pounds at times.