The Utah Education Initiative — Education Mentor Chat is a training and knowledge service that supports the Education Mentor calling. The solution employs a custom Gemini-based model grounded in an authorized corpus of program materials, Google Drive documents, official guides, and transcripts from training webinars/Zoom meetings. Education Mentors may submit natural-language questions (e.g., “What are the top five priorities for this calling?”) and receive concise, policy-aligned responses with references to the underlying sources. The tool is designed to support onboarding and ongoing proficiency while promoting consistency of guidance across the program.
Accelerate onboarding and ongoing training for Education Mentors by delivering fast, accurate answers drawn from approved resources.
Increase consistency of guidance by grounding answers in the same curated knowledge base used across the program.
Save time by reducing the need to search or watch long recordings; mentors can ask targeted questions and get actionable summaries.
Enhance mentor effectiveness as they assist individuals in exploring higher education and upskilling opportunities.
Curated Knowledge Base
The model is trained on an official Google Drive collection related to the Education Mentor calling, as well as text transcripts generated from past training webinars/Zoom meetings.
AI Retrieval & Answering
When an Education Mentor asks a question, the system searches those sources and produces a concise, plain-language answer; when possible, it points back to the underlying document or training segment.
Iterative Improvement
The team is evaluating answer quality and refining prompts, sources, and guardrails based on early testing. Platform and access methods are being finalized (Gemini today; final delivery channel TBD).
Who It’s For & How It Would Benefit Them
Primary audience: Utah Education Initiative Education Mentors at ward and stake levels.
Benefits to mentors: Faster learning curve, on-demand answers, and consistent guidance while balancing calling responsibilities with everyday life.
Whom mentors support: Primarily members, but mentors may also assist non-members as they connect people with education and skills pathways.
Progress & Next Steps
Current status
The solution has moved from internal prototyping to a limited pilot. A custom Gemini-based assistant has been trained on the approved Google Drive corpus and on transcripts from training webinars/Zoom meetings, and it has now been shared with a selected group of Education Mentors/stakeholders for hands-on testing. The assistant is also now working in both English and Spanish, so mentors can interact with it in either language.
What this means
We are no longer only testing internally; we are collecting feedback from real users in the target role.
We can validate answer quality, tone, and source coverage in two languages.
We can identify gaps in the central repository (missing topics, outdated guidance, or unclear instructions) and update it centrally.
Immediate next steps
Collect and review pilot feedback on accuracy, usefulness, clarity of answers, and EN/ES behavior.
Refine the training corpus (Google Drive materials + meeting/transcript content) and improve prompt design based on that feedback.
Confirm the delivery channel and authentication approach (whether to remain in the current Gemini environment or surface it in another user-facing channel).
Define an update and governance cadence so new or revised Education Mentor materials can be added by the project team and immediately reflected in the assistant’s responses.
Prepare for broader availability once the pilot shows consistent, policy-aligned answers.
1) What is the Education Mentor Chat?
An AI-assisted Q&A tool trained on official Education Mentor materials (documents plus transcripts from training meetings) that answers calling-related questions in seconds.
2) Who can use it?
Education Mentors in the Utah Education Initiative (ward/stake). Access for a wider audience will be considered after pilot learnings.
3) Is it approved and available today?
Not yet. It’s an early prototype undergoing internal review; pilot timing and the final delivery platform are still being determined.
4) Where does the information come from?
A curated Google Drive repository and text transcripts generated from recorded training webinars/Zoom sessions. Answers are grounded in these sources.
5) Will answers include references?
When possible, the tool indicates which resource or training segment supports an answer so mentors can verify details quickly.
6) Does it replace official training?
No. It complements existing materials by making them searchable and summarized; mentors should continue to follow official guidance and training.
7) What about accuracy and limitations?
Because it’s early-stage, answers are being evaluated, and the knowledge base is being refined. Users should verify critical details in the linked source material.
8) Will it expand beyond Utah?
The calling is in pilot; broader expansion may be considered after Utah learnings, subject to approvals.
9) Do Education Mentors need to “train” the chatbot for it to work?
No. The assistant is pre-trained by the project team on an authorized corpus of Education Mentor materials (Google Drive documents, official guides, and transcripts from training webinars/Zoom meetings). Mentors simply ask questions in natural language; the system retrieves from those sources and returns concise, policy-aligned answers with references when available. Ongoing curation, model updates, and governance are handled by the project team—not by end users. If gaps are identified, mentors may provide feedback, and the team will update the underlying sources as appropriate; mentors should not upload or input personal or sensitive data.