PDP Assist
What It Is
A side-project demo of an AI learning coach for junior engineers. You pick a career journey, rate your skills, answer a few coach questions, and get a pathway built from real course modules — then put the training on a calendar.
The learning content is aimed at Azure skills. The coach uses Google Vertex Gemini when it is set up; if not, a local fallback still runs the demo. There is no sign-in. Progress is saved in the browser. Use Reset demo if you want to start over.
Live at pdp-assist.vercel.app.
The Interesting Bit
You curate a library of training. The coach then builds a plan from that library, tailored to the individual developer — their journey, skill gaps, and how much time they can give. Same catalogue; different pathway for each person.
Try It
Next Steps
The demo is rough on purpose. Natural next steps:
- From demo to workplace tool — real sign-in and saved accounts, so a personal development plan survives more than one browser.
- Capability frameworks — map skills to something like the DDaT Capability Framework so pathways can support promotion evidence, not just a reading list.
- A wider catalogue — plug in Microsoft Learn, Pluralsight, or internal courses, while still only picking real catalogue IDs.
- A closed loop — finish sessions, reassess skills, refresh the pathway. Optional views for managers or L&D for check-ins and cohorts.
- Calendar outside the tab — export to ICS, sync with Outlook or Google, Slack reminders before a study block.
- Nearby use cases — graduate programmes, contractor onboarding, career switches (analyst to engineer), anonymised skill-gap views across a team.
Static learning catalogues are easy. Sticking to a plan is the hard part — that is the gap this is aiming at.
Between “here’s a learning portal” and “I actually did the modules” is a lot of empty calendar.