Employee Offloading Intelligence

Inspiration

When employees leave, companies recover their laptops and revoke access—but often lose years of undocumented decisions, workarounds, and relationships. I wanted to make knowledge transfer proactive, targeted, and useful before that expertise walks out the door.

What it does

Employee Offboarding Intelligence analyzes a departing employee’s work artifacts to identify knowledge only they may hold. It highlights risks with supporting evidence, generates focused exit-interview questions, and turns the answers into searchable, cited company knowledge.

How we built it

I built a Next.js and TypeScript interface with shared, type-safe data models. Triggerware orchestrates the knowledge-risk scanning step, with Zod validating agent output and a deterministic fallback keeping the prototype reliable. The broader platform uses NestJS, PostgreSQL, Prisma, Redis, and Tailwind CSS.

Challenges we ran into

The hardest problem was distinguishing missing knowledge from merely scattered knowledge. We also had to present AI findings transparently, avoid unsupported answers, and keep the experience functional when the external analysis agent was unavailable.

Accomplishments that we're proud of

We created a complete before-and-after workflow: detect a knowledge gap, show the evidence, conduct a targeted interview, preserve the response, and demonstrate that the company can now answer a previously unanswerable question.

What we learned

I learned that offboarding should capture reasoning, not just files and tasks. I also learned that trustworthy AI needs citations, clear uncertainty, validated outputs, and dependable fallbacks.

What's next for Employee Offboarding Intelligence

Next, I plan to connect real workplace sources such as Slack, email, ticketing systems, and document platforms; expand Triggerware across the full workflow; add human review and privacy controls; and track knowledge risk continuously—not only after someone gives notice.