Asha Abdulkadir

02 · Applied AI · Ongoing

Applied Artificial Intelligence.

Applied AI runs through both halves of my work: the systems I ship at an enterprise service platform, and the graduate study I am doing at Lipscomb University. This section holds both.

No. 01

Featured projects

Applied AI I have put into production, and the builds I am working through outside of work.

Knowledge surfacing and triage with Rovo AI

Shipped · Enterprise service platform

Integrated Rovo AI into an enterprise service platform to surface relevant knowledge articles and triage incoming work, which lifted both the speed and the consistency of first response. The interesting part was not the model. It was the knowledge architecture underneath it: retrieval is only as good as the structure and permissioning of what it retrieves, so the Confluence taxonomy and the article templates had to be designed first for the AI layer to be worth anything.

Read the full case study, filed under Enterprise Systems & Automation →

No. 02

Research & coursework

I am reading for a Master of Science in Applied Artificial Intelligence at Lipscomb University, taken part time alongside full time engineering work. The programme is the formal half of how I approach AI; the service platform is where the ideas get tested against real users and real constraints.

What I am most interested in is the unglamorous layer: how enterprise data has to be structured, governed, and permissioned before an AI feature on top of it is trustworthy. That question connects directly to the pipeline and knowledge architecture work elsewhere in this portfolio.

No. 03

Reading & ideas

The questions I am actively working through, in the programme and at work:

  • Applied AI in IT service management

    Where assistive AI genuinely reduces load, and where it just moves the work

  • Retrieval augmented generation in enterprise contexts

    Why retrieval quality is a data architecture problem before it is a model problem

  • Responsible AI deployment

    What has to be true before a system is allowed to act rather than suggest

  • AI governance for midsize organizations

    Governance that fits teams without a dedicated risk function

← Back to projects