Taylor

User Research & Product Development

2026

Research, Testing, Marketing

Collaborated with product, development, design teams, and our CTO to create, test, and develop CareerBeacon's AI Job Search Copilot, Taylor.

Team objective: Develop a minimum viable product within a 2-month sprint, moving from internal testing to a pilot launch of 1,000 users to evaluate usability and interaction.

My objective: As the lead on marketing and user experience, I focused on developing Taylor’s values, character guidelines, and knowledge of the Canadian jobseeker experience to ensure the tool felt emotionally resonant and trustworthy rather than purely transactional.

I had the following questions during this project:

1. What are the most acute pain points for Canadian jobseekers, and how can Taylor solve them while making users feel heard and understood?
2. What psychological framework does an AI tool need to build rapport with human beings without straying from its core functional tasks?
3. How do users of different backgrounds (ethnicity, language, location) interact with AI, and how should Taylor adapt its tone to minimize friction?
4. Can foundational emotional intellgence be built into the tool to build long-term trust with a potentially vulnerable user base?

To ensure Taylor was grounded and understood Canadian jobseeker needs, I led several research and testing initiatives:

1. Automated Sentiment Research: Helped create an AI automation to analyze hundreds of Reddit threads daily, identifying common language, emotional intensity, and market trends impacting the Canadian jobseeker experience.

2. Emotional Journey Mapping: Mapped the jobseeker's emotional arc to determine the right tone for Taylor to adopt at different stages of the job search journey.

3. Stress Testing and Evaluation: Conducted extensive adversarial testing by prompting Taylor with messages varying in intensity, language, and grammar. I also tested actual LinkedIn messages we’ve received from jobseekers to hopefully mimic external usage.

4. Prompt Development: Developed some of Taylor’s core identity components, including personality traits, tone of voice, background history, and values.

5. Initial User Conversion: Wrote the initial outreach email from Taylor, prioritizing an emotionally resonant hook to grab the reader’s attention and make them want to engage with Taylor.

Taylor is still in external testing, but here's what we've learned:

Engagement Metrics: The initial welcome email achieved a 54% open rate.

Iterative Feedback Loop: We currently monitor and label every response via a colour-coded evaluation system (Strong / Inconsistent / Needs Improvement) to document what we learn and refine Taylor’s prompting.

Current Focus: Reducing onboarding friction by A/B testing a more focused version of the outreach email, as well as building familiarity with Taylor amongst CareerBeacon’s existing users by introducing Taylor in other high-traffic areas of our site. We are also adjusting Taylor's prompt to give it more emotional intelligence and are testing it using a Lovable prototype experience picture below.