Stella Rolfson

Why this persona fits the audience

Builds client chatbots and document workflows for startups across London and Manchester. She estimates token usage for each proposal, pastes sample prompts into calculators, and documents official-model versus gateway pricing in her project notes.

• Age: 34
• Location: London, UK
• Occupation: Freelance AI automation consultant
• Income: £110k equivalent
• Education: BSc Information Systems
• Company size: solo practitioner
• Years in role: 7

“I didn't leave a 9-to-5 to become a glorified API pricing gatekeeper, but here we are.”

Stella Rolfson's journey to freelance consulting began in Manchester, where she earned her Information Systems degree and took her first job maintaining CRM systems for an HR startup. Her big break came when she automated a client's painfully manual onboarding process using early LLMs, reducing their time by 80%. The experience hooked her on AI automation, and in 2017, she went solo, building chatbots and document workflows for London and Manchester's burgeoning tech scene. Her meticulous approach to cost estimation—a mix of spreadsheets, custom calculators, and extensive testing—became her signature. The current market surge has brought new challenges: clients expecting AI miracles, ever-shifting API pricing, and the constant balancing act between quality work and profit margins. Recently, she turned down a lucrative but vague project from a US-based edtech startup, sticking to her policy of requiring clear token usage boundaries upfront. Most evenings, she's found updating her internal pricing comparison documents or answering last-minute client questions. Weekends might involve outside commitments: a startup accelerator panel or teaching a workshop on responsible AI implementation. She's currently considering whether to develop a standardized workflow template, but the idea of finding 20 hours to build it while keeping her client commitments feels impossible. What keeps her going? The rare client testimonial about how her work changed their business—and the fact that her sister finally stopped asking when she'd get a 'real' job.

Background & daily life

A day in their life

6:30am: Matcha and a quick scroll through Hacker News. 7:15am: First drafting client emails about yesterday's testing results. 8:30am: Client call with a Manchester fintech startup panicking about their token budget. 11:00am: Coding a custom workflow automation for a client who 'just needs a simple solution' (it never is). 1:00pm: Lunch and a 15-minute walk to clear her head. 2:30pm: Reviewing API pricing changes and updating her internal comparison documents. 4:00pm: Debugging why a client's test chatbot is generating gibberish. 6:30pm: Slack messages from a client about 'one quick question' that turns into an hour-long conversation. 8:00pm: Yoga or cycling, depending on how much she's been sitting. 9:30pm: Quick client call with a US-based startup (timezone differences are brutal). 11:00pm: Final review of tomorrow's schedule before bed.

Lifestyle

Stella maintains a meticulously organized one-bedroom in Peckham, where every surface serves dual purpose. Her morning routine includes making matcha (never coffee) in a mug labeled 'Do NOT Use for Client Work' and reviewing her Trello dashboard. Weekends involve cycling along the Thames Path or volunteering at a startup accelerator. She budgets obsessively but splurges on ergonomic office gear. Socially, she rotates between a tight-knit group of freelancers and her university network, though she's been avoiding her family's constant 'when are you getting married' texts. Health-wise, she's trying to get back to yoga after a series of late-night debugging sessions ruined her shoulder.

Work life

Stella's work revolves around spreadsheets and Slack DMs. Typical days start with reviewing client workbooks tracking token usage—she's currently helping an edtech startup optimize their LLM costs after their last bill shocked them. Tools of choice: Notion for project tracking, custom Python scripts for cost prediction, and an ever-growing folder of .txt files with sample prompts she's tested. Her biggest stressors are scope creep and clients who assume 'AI' means zero maintenence. She recently turned down a lucrative but vague project because the client couldn't articulate their token usage limits. Workday rhythm: Deep focus mornings, client calls after lunch, and evenings spent answering urgent Slack messages while meal-prepping.

Relationships

Stella's closest confidants are her university study group, now scattered across the tech world, who she meets weekly for virtual happy hours. She maintains professional friendships with former colleagues from two startups, but her strongest bond is with her older sister in Edinburgh—the only family member who truly understands her work. Her parents' constant questions about 'real jobs' grate on her, but she indulges them during annual Christmas calls. In London, she's part of an informal freelancer network that meets at Hackney coffee shops, though she rarely attends—her calendar is too tightly packed with client onboarding sessions.

How they decide

Methodically over-researches before committing, but once decided moves quickly. Maintains spreadsheets of pros/cons and waits 48 hours before final decisions. Trusts but verifies vendor claims, and asks for exact costs in pricing discussions. Prefers written communication for complex topics, though she'll talk through decisions with her university friends.

How they sound

Speaks in complete sentences but with frequent asides like 'you'll appreciate this when you see the token count'. Uses Britishisms ('biscuits', 'holiday') but switches to American terms in tech contexts. Avoids jargon unless explaining concepts and then becomes unexpectedly patient. Will apologetically interrupt herself with corrections. Has a habit of ending sentences with 'does that make sense?' when working through technical concepts with clients. Occasionally makes dry observations about human behavior in meetings ('this is the third time this week I've explained that LLMs don't have opinions').