I’m Nandita, a machine learning engineer and AI risk manager. I lead teams building AI systems across healthcare, biodefense, and research—focused on infrastructure that preserves provenance, enables reproducibility, and connects model behavior to governed data.
If your model development stack values speed, control, and auditability—I’d love to connect.
I’m a connector—someone who brings together teams, systems, and goals to make AI work in the real world. To me, responsible AI comes down to three things: well-governed data, clear roles, and infrastructure that scales. With experience in both ML engineering and strategy, I help bridge technical work with broader organizational goals. Whether I’m working with developers, data stewards, or leadership, I focus on building AI pipelines that are accountable, aligned, and built to last.
Directed technical strategy for the NIH Common Fund Data Ecosystem (CFDE), a federated $50M biomedical data platform spanning 12 NIH centers. Authored NIH’s first Data & AI Strategy, aligning with EO 13960 and the NIST AI RMF to define AI governance implementation measures. Developed AI readiness rubrics integrated with existing grant evaluation frameworks. Led proposal down-selection and authored pay plans for AI and data awardees, totaling $7.2M. Co-led an NIH-wide task force to standardize procurement milestones for AI- and data-enabled research. Promoted low-code platforms and LLM tools to scale responsible AI literacy across scientific teams.
Launched the program’s first enterprise-wide Data Governance Committee and stewardship program across 8 domains (EHR, genomics, surveys, etc.) for a 650,000-participant cohort. Led modernization efforts using Palantir Foundry, delivering an AI audit toolkit and data quality scorecards. Co-developed a privacy-preserving record linkage (PPRL) approach adopted for multi-institutional data use. Increased operational efficiency by 30% through a unified Data & AI strategy, and hosted responsible AI hackathons to operationalize model transparency.
Advised federal clients on AI governance, auditability, and ethical deployment. Developed model risk scoring tools in Deloitte’s internal GenAI sandbox, improving performance on high-stakes use cases by 20%. Co-authored the HHS Trustworthy AI Playbook and secured $15.7M in competitive contracts. Designed Deloitte’s first Responsible AI course and co-created a capstone at UC San Diego on explainability (SHAP, LIME, fairness dashboards). Contributed policy recommendations to EO 13960 and the NIST AI RMF.
Developed NLP algorithms to detect adversarial bioengineering threats under IARPA's FELIX program, achieving >90% accuracy in evaluation. Built ML and bioinformatics pipelines in R, Python, and SQL. Led a $150K seed-funded internal innovation team and reduced genomic data processing costs by $200K. Produced reproducible research documentation using Atlassian tools and collaborated with federal partners including the FBI.
Built high-throughput genome sequencing pipelines to support clinical vaccine trials. Analyzed single-cell and bulk RNA-seq datasets to uncover immune correlates of protection. Deployed SLURM-based compute workflows on HPC environments and authored reproducible R/Python analysis code.
Led FDA 510(k) regulatory submissions for Class II diagnostic devices, including KRAS, BRAF, and CYP2C19 genotyping panels. Served as Biological Safety Officer and developed lab infrastructure, protocols, and safety training for BSL-2 operations. Published a peer-reviewed qPCR assay (T-blocker) capable of detecting somatic mutations down to 0.1% frequency.
Co-developed and taught an undergraduate capstone on responsible AI, with technical labs using SHAP, LIME, and model fairness dashboards. Designed hybrid curriculum focused on auditing real-world ML systems.
Focused on bioinformatics, machine learning applications in genomics, and statistical modeling for biological data. Thesis work on regulatory network analysis and computational approaches to understanding gene expression patterns.
Comprehensive foundation in molecular biology, genetics, and biochemistry. Emphasis on quantitative analysis and research methodology preparation for graduate studies in computational biology.
From Sandbox to Strategy: Full-stack Skills for Scaling ML Systems Responsibly
Interested in collaborating on AI risk strategy or ML infrastructure? I’m currently open to new opportunities.