Citi
Project Manager (contract) working on the response and explanation layer for a credit-card fraud scoring model, alongside fraud and risk teams.
Read the Citi case studyMadan in 60 seconds
This is the plain brief: the same evidence, without the airport. Project Manager at Citi · PM / TPM · AI delivery. Before that, Product Analyst at NaviSite. Seattle, CAPM certified. I coordinate Product, Engineering, ML, and QA to make complex initiatives executable.
Ants Corp
Software, cloud migration, cybersecurity, and SQL-based UAT. Standardized intake reduced onboarding from two weeks to eight days.
Read the intake case studyChoose the work most relevant to you.
Review project capabilities here. Repository and live-demo links below are optional and open a new tab.
Sentinel Eval Harness
Run cases, calculate pass rates, compare baselines, and enforce configured release thresholds.
AI Project Manager Assistant
Generate structured tasks, dependencies, assignments, milestones, and RAID items from a brief and roster.
AI Experimentation Platform
Save drafts locally, plan samples, enter conversion counts, check guardrails, and record decisions.
Sentinel Eval Harness · implementation evidence
Repeatable evaluations and release gates
Run cases, calculate pass rates, compare baselines, and enforce configured release thresholds.
Scope: External model adapters need their own endpoint; source inspection does not establish production performance.
Inspect the implemented files
- src/eval_harness/runner.py · GitHub, new tab
- src/eval_harness/scorers/deterministic.py · GitHub, new tab
- .github/workflows/evaluation-gate.yml · GitHub, new tab
- tests/test_harness.py · GitHub, new tab
Files inspected October 8, 2026. This confirms implementation is present; it does not certify every deployment or external integration.
Match a role to the evidence.
A transparent keyword map, not an automated hiring score. Text stays in this page and is not sent to a service.