AI products
Practical AI, managed risk
An internal AI assistant that made everyday work easier without treating sensitive information as an afterthought.
Try sending the sample message below.
Proceed with the email redacted?
- My role
- Discovery, priorities, security coordination and rollout.
- Context
- Employee-facing application
Useful AI, sensitive data
Employees needed help summarizing information, researching regulatory questions and drafting emails. We built a general-purpose chat application with controls for personally identifiable information and confidential client data. Firm-wide licensing costs helped make the case for an internal tool.
Product leadership
I led the product work from discovery through rollout, bringing employees, business stakeholders and security into the decisions. Engineering built the application.
Guardrails, not blockers
The hardest tradeoff was adding safeguards without making the tool frustrating or limiting its usefulness. Controls had to fit a regulated workplace while preserving the smooth, feature-rich experience employees expected.
Adoption and value
Approximately 700 employees had access; roughly half used it weekly during the first three months, based on internal admin reporting. Feedback highlighted the perceived accuracy of regulatory research using regulatory sources and web search, and the value of an internal alternative to paid tools.
These are adoption and feedback observations, not measured cost savings or a reduction in security incidents.
Let’s talk.
Have a role or a project in mind? Tell me what you’re working on.