I didn’t start with a framework or a warning about disruption. I started with excitement.
“Look, if you’re graduating into AI right now, you’re walking into the most opportunity-rich market I’ve seen in years.”
That’s how I opened my session at Thompson Rivers University. I told the room of Computer Science and Engineering students that I work in one of the most energizing corners of the field: avatars, because it blends engineering with storytelling, presence with practicality. I wasn’t there to predict the future; I was there to demonstrate how a human double can carry your voice, how an interactive coach can help you practice interviews, and how these tools translate into real projects, internships, and startups.
Made possible by the Discovery Foundation’s AI Skills Accelerator
This session was part of the Discovery Foundation’s AI Skills Accelerator, delivered by the Central Interior Business Accelerator (formerly Kamloops Innovation) and TRU’s Computing Science and Engineering departments.
It was a generous, high-trust day where multiple presenters volunteered time to help students build not just technical skills, but the people skills that carry you through interviews, internships, and the messy middle of new ventures.
I was invited to focus on avatars — what they are, how they’re evolving, and why they’re quickly moving from novelty to necessity. The Accelerator is structured around workshops, mentorship, and student engagement, and it’s designed to connect AI learning with real-world outcomes.
But if there’s one takeaway I want readers to remember, it’s this:
Avatars don’t replace human skills — they demand them.
From talking heads to teammates
When most people hear “avatar,” they picture a talking head that reads a script. That’s certainly one mode, but it’s only one face of a broader progression I laid out for the group:
Synthetic avatars: brand-safe personas with no real-world twin, great for 24/7 explainers, onboarding, and FAQs.
Human doubles (scripted): your own likeness — my “Virtual John” — delivering consistent micro-lessons, personalized welcomes, or quick thank-yous at scale.
Interactive avatars: two-way, adaptive conversations that respond to context and nuance in real time — closer to a teammate than a teleprompter.
That third category is where the ground is moving the fastest. It’s also where the soft skills matter most.
An interactive avatar isn’t just “content delivery.” It’s a dance about listening, asking follow-ups, moderating tone, and deciding when to push or pause. Under the hood, these systems work best when we “train them like an executive assistant” — not just with facts, but with values, voice, guardrails, and judgment.
Think about that for a second. To make an avatar more helpful, we have to get clearer about our own priorities and principles.
Tools are making us do our homework.
Why voice is still the make-or-break
I also told the students something I’ve learned the hard way: voice carries trust.
The same words spoken with different vocal styles can land as a lecture, a conversation, or a supportive nudge. I demonstrated four voice profiles I use — Narration, Presentation, Conversational, and Supportive — and how we capture and protect those recordings.
The point wasn’t technical; it was human. When your voice shows up in multiple contexts (a live keynote, a micro-lesson, an avatar’s answer), consistency becomes a proxy for credibility.
If you want your future interactive agents to reflect you, invest in your voice now: practice clarity, record cleanly, document your tone, and set rules for where your voice can and cannot be used. It’s not glamorous, but it’s the kind of foundation that prevents reputation debt later.
The interview coach named “Rez-U-May”
Because this workshop had a strong “people skills” theme — focused on how to interview, communicate, and work on teams — I built something special for the day: an interactive avatar that acts as an interview coach.
TRU Professor Kevin O’Neil’s wife came up with the “play on words” name: Rez-U-May.
I trained Rez-U-May with the same material I teach on how to structure concise answers, when to offer examples, how to handle “Tell me about a time…” prompts, and how to close an interview with genuine curiosity.
During the session I did a live demo: Rez-U-May greeted me, asked for the role I was preparing for, and immediately adapted its questions to match the scenario. Students were sent the link so they could practice their own answers.
Here’s the important part: Rez-U-May doesn’t give anyone a script to memorize. It gives feedback.
Shorter story. Clearer result. Stronger evidence.
That’s the difference between automation and augmentation. We’re not outsourcing the interview — we are practicing it.
Practical learnings from the session
A few themes stood out from the questions and conversations in the room:
Students are hungry for applied clarity.
Many are in their fourth year, looking at a job market that has changed while they were studying. They are looking for a concrete workflow — like how a human double can personalize outreach, how an interactive avatar can rehearse feedback. Tools become plans. Plans become momentum.Soft skills are the leverage point.
The higher the capability of the tool, the more it rewards human judgment. We see this in every domain: you get more out of an avatar when your prompts are specific; you get more out of an interview coach when you’ve reflected on your own stories; you get more out of a human double when you’ve articulated your values and voice.Trust comes from context.
The same avatar can be a delight in one scenario and a disaster in another. We talked about consent, disclosure, and expectations. When to tell people they’re talking to an avatar (almost always), how to escalate to a human gracefully, and where to set boundaries on data retention and usage.
The students asked sharp questions, and I was glad they did. That curiosity is a career-long asset.
Mentorship matters (and scales the right way)
One of the best parts of this Discovery Foundation program is that it doesn’t stop at workshops. It funds time for one-on-one mentorship, and I’ve been fortunate this semester to work with three students — Mose Tule, Mohammed Abbas, and Madhurima Sen — on their startup ideas and next steps.
Mentorship turns a tool demo into a habit. It creates a feedback loop, and when you pair that with a community, faculty who open doors, an accelerator that connects dots, and a foundation that insists on outcomes, you get progress that shows up in the real world, not just in a slide deck.
The Discovery Foundation’s model includes a cadence of workshops, labs, and coaching across the year, with mid-term and final reporting to keep everyone honest about impact. That attention to measurable outcomes is part of why I’m proud to contribute.
But should we really be doing this?
We also talked candidly about the “should.”
Should we train avatars and agents to mimic us?
Should we put a human double into a sales nurture flow?
Should we let an interview coach listen to practice answers?
I offered my current stance, which is less a conclusion and more a compass:
Disclose clearly. If someone is talking to an agent, tell them. Let consent and context lead.
Design for escalation. “I can transfer you to a human now” is a feature, not a failure.
Protect the master. Keep original voice recordings secure, ensure revocation rights, and be explicit about where your voice can’t go.
Coach for judgment. Whether it’s a student or an avatar, the heart of the work is still prioritization, empathy, and honest trade-offs.
It’s not a perfect map, but it keeps me oriented toward trust.
What avatars teach us about being human
If that sounds lofty, let me ground it.
Teaching avatars has made me more attentive to human signals, not less. When I build a training set, I have to write the words I actually want to say, in the tone I actually want to use, with the values I actually want to hold.
That reflection becomes a mirror. And weirdly, once you’ve documented your voice and judgment for a machine, you tend to show up with more consistency as a person.
That’s the irony and the gift of this work: we design better avatars by becoming better communicators.
Gratitude (and why this program exists)
None of this happens in a vacuum.
The Discovery Foundation’s Technology Education Program invests in applied learning workshops, mentorship, and hands-on student engagement — with an aim at measurable outcomes across the year.
The Central Interior Business Accelerator and TRU Computer Science and Engineering departments are the delivery partners that get rooms booked, students recruited, and speakers organized, with reporting and accountability built in.
It’s a collaborative stack that moves resources where they matter most: into communities, classrooms, and small businesses who need practical help to adopt AI.
So, thank you to the Foundation for the support, to TRU faculty and staff for the warm welcome, and to the other presenters who shared their time and craft with students. Days like this remind me that “ecosystem” isn’t a buzzword; it’s people choosing to do the slow, generous work of building capacity together.
The future these students will enter
If you entered university four years ago dreaming about a narrow job description, you’re graduating into a wider field. The next few years will reward people who can do three things at once:
Think with agents and avatars (design workflows, set guardrails, evaluate results).
Speak with clarity (in meetings, in writing, through human doubles).
Practice with feedback (from mentors, peers, and yes, interview coaches like Rez-U-May).
Avatars aren’t replacing any of that. They’re creating new rooms where those skills can be practiced more often, with less friction, and with fewer gatekeepers.
As for me? I’ll keep building, teaching, and mentoring in this space because every time I help a student find their voice, I find a little more of mine.