AI Skills Academy — Online Education and Digital Innovation

AI tools are reshaping how we work, learn, and create. When we asked ourselves how to democratize AI upskilling and literacy across the University community, we landed on a simple premise: meet learners where they are, on their schedules.

From December 10, 2025 to January 10, 2026, UVA Online Education and Digital Innovation (OEDI) partnered with leading skills assessment platform CodeSignal to deliver our first-ever AI Skills Academy. Through no-cost, asynchronous AI upskilling, 50 UVA students and 75 faculty and staff increased AI literacy and confidence across several core competencies.

Program Structure & Delivery

Over a 40-day access period, 175 participants engaged in fully asynchronous, self-paced AI literacy upskilling coursework. Based on experience level, participants chose between two learning tracks:

A diagram featuring two corporate minimalist-style icons discusses two learning tracks - AI use and Business Applications of AI.

This visual breaks down both learning tracks in the AI Skills Academy. The first, AI Use, allowed users to explore AI fundamentals such as prompting basics, image and video generation, and AI ethics and limitations and learn more about emerging LLMs and use cases. The second, entitled Business Applications of AI, allowed users to explore more advanced topics such as retrieval-augmented generation (RAG), advanced prompt engineering, and chatbot creation.

As a supplement to virtual coursework, participants were invited to participate in several synchronous opportunities to share insights and troubleshoot concerns. These included:

An optional kickoff webinar on December 3, 2025
Pre- and post-learning surveys to measure growth and identify areas of interest
Weekly virtual office hours
In-person retrospectives at our office to share real-time feedback and compare perspectives

This structure ensured that participants could learn on their own time while having a breadth of resources and engagement opportunities at their disposal.

Engagement Data

User Activity Comparison

Among all participants, 64% of faculty/staff and 46% of students were active users. "Active" users are defined as those who logged into CodeSignal and completed at least one practice exercise.

Two segmented bar graphs highlight the proportion of faculty/staff and student users listed as active (hot pink) or inactive (navy blue) on a platform

A segmented bar graph displays two cohorts on the X axis - students and faculty/staff. The Y axis indicates the number of users, with segments in navy blue and pink marking the amount of inactive and active users, respectively. Out of 50 student users, the number of active users rests at 23, with 27 inactive members. Out of 75 faculty/staff users, 48 users are listed as active and 27 as inactive.

Depth of User Engagement

Faculty/staff users spent more days on average learning on the platform, signaling a more sustained engagement with CodeSignal’s content and commitment to learning beyond surface-level exploration.

Two sets of bar graphs display user activity by day for two user cohorts (faculty/staff and students).

A bar chart displays student and faculty/staff cohorts on the X axis and the number of users on the Y axis. Colored segments display all users (hot pink), active users (navy blue), users active for five or more days (blue), and users active for ten or more days (black). Out of 50 student users, 23 are listed as active, 7 as active for 5+ days, and 2 as active for 10+ days. Out of 75 faculty users, 48 are listed as active, 16 are listed as active for 5+ days, and 4 as active for 10+ days.

The table above displays activity among users. An "active" day is defined as a day in which a user logged into CodeSignal and began or completed coursework.
Metric	% of Student Users	% of Faculty/Staff Users
Active 5+ days (% of all users)	14%	21%
Active 10+ days (% of all users)	4%	11%
Average active days	3.6	6.9

Practice Completion Intensity

Higher rates of platform engagement aligned with practice exercise completion and signaled more consistent learning among the faculty/staff cohort.

Two funnel charts display a breakdown of user activity by practice exercise completion among two different cohorts - students and faculty/staff.

Two funnel charts, one for students and one for faculty/staff, break down practice exercise completion among each cohort. Hot pink represents active users, navy blue represents those who completed at least 25 practice exercises, and blue represents those who completed at least 50 practice exercises. Out of 75 faculty/staff users, 32 completed at least 25 practices, 21 completed at least 50 practices, and four completed 100+ practices. Out of 50 student users, 14 completed at least 25 practices, 9 completed at least 50 practices, and 3 completed 100+ practices.

The table above breaks down practice completion by milestone and user type, with faculty/staff users showing generally higher completion rates than student users.
Milestone	Student Users	Faculty/Staff
25+ Practices Completed	28% (14/50)	42.7% (32/75)
50+ Practices Completed	18% (9/50)	28% (21/75)
100+ Practices	6% (3/50)	5.3% (4/75)

The "Super User" Phenomenon

While practice completion varied between student and faculty/staff cohorts, both produced a similar number of "super users"--defined as users who completed at least 100 practice exercises over the program period.

Key Insights

Confidence Gains Across the Board

Participants reported notable confidence gains in key AI skills over the 40-day period, with the strongest growth emerging in the following core competencies:

Prompting skills: +25% confidence gain (3.41 → 4.25/5)
Personal applications of AI: +28% confidence gain (2.92 → 3.75/5)
Awareness of AI ethics & limitations: +20% confidence gain (3.37 → 4.05/5)

A bar graph compares user confidence in AI competencies pre-project (navy blue) and post-project (hot pink).

A bar chart displays levels of confidence reported by all learners both before and after the project period. Navy blue represents pre-project confidence ratings, and hot pink represents post-project ratings. Skills (prompting confidence, work implementation, personal use, ethics & limitations, and general comfort) are listed on the X axis, and ratings from 1-5 are listed on the Y axis. Across all categories, confidence levels rise.

Break Timing Works

While it can be tempting to believe that faculty, staff, and students generally dislike learning outside of scheduled academic term, learner feedback overwhelmingly favored delivery over scheduled academic and professional breaks.

Professional Audiences are Ready

The 64% faculty/staff participation rate, along with sustained engagement during outreach and wrap-up phases, signals strong demand for AI upskilling among professional audiences.

Platform Feedback

While CodeSignal effectively built foundational skills, learners identified opportunities for improvement. Some learners preferred a more regimented and less gamified format, while others found the first track too easy and the second too difficult.

Learner Feedback

"I didn't realize prompting was such a useful skill...it has saved me time when I need the LLM to respond accurately the first try."

- Second-year student

"I see more use cases for AI than I did previously, especially knowing how to better set up the context and provide constraints."

- UVA staff member

"I plan to use the skills I've learned to create better prompts and create more refined responses."

- Third-year student

"I found [CodeSignal] enjoyable and a little bit addictive in the best way. I liked doing this over break."

- UVA faculty member

What's Next

The AI Skills Academy demonstrated that accessible, well-timed AI literacy programming can drive meaningful engagement and confidence growth across diverse university audiences. We seek to build on these insights as we scale professional development opportunities, both for UVA students and UVA faculty and staff.

Have questions about the program? Interested in participating in future learning pilots?

Share your thoughts with us by filling out this form or contacting oedi@virginia.edu.