10 fictional AI personas. Exploratory perspectives, not a survey of real customers. Methodology & limitations

Sjors Willems

Sjors Willems

Product founder / analyst at a small product studio

Full-depth AI personas

Why this persona fits the audience

Is the founder of a 9-person Amsterdam product studio that is testing an AI support agent for a client. He writes lightweight scripts to group conversations, compares model runs in a shared dashboard, and documents findings in Notion for the account lead.

  • Age: 31
  • Location: Amsterdam, Netherlands
  • Occupation: Founder and product analyst
  • Income: €68k
  • Education: MSc Human-Computer Interaction
  • Company size: 9 employees
  • Years in role: 3
“I don't mind being the smartest person in the room, but I'd rather be the one who asks the right questions about what we're actually building for people.”

Sjors Willems grew up in a small town outside Utrecht, the son of an engineer father and a teacher mother. His early fascination with how people interact with technology led him to study Human-Computer Interaction at TU Delft, where he specialized in conversational interfaces. After graduating, he worked briefly at a Berlin startup before moving back to Amsterdam to found his own product studio. The 9-person team he built now focuses on AI implementations for mid-sized companies, with a focus on support and customer service applications. His current project involves testing an AI agent for a German SaaS client, where he's using Python to group and analyze conversation data that gets visualizes in a shared dashboard. He's a lean founder - their small budget means they often build their own tools rather than buying off-the-shelf solutions, which Sjors both appreciates and finds frustrating when time is tight. His studio operates on a hybrid model, with most of the team working remotely, which he finds challenging for spontaneous collaboration but necessary for attracting talent. Financially, he prioritizes profitability over growth, keeping costs low by doing much of the technical work himself. He's particularly proud of their recent decision to move all operations to green hosting providers, despite the higher costs. Professionally, his biggest constraint is managing client expectations - many are new to AI and expect quick, flashy results while he insists on careful, iterative development. Personally, he's been trying to improve his public speaking to better communicate complex technical concepts to non-technical stakeholders. His social circle is small but tight - he spends most of his evenings either working on open-source projects or repairing his increasingly complex bicycle (he recently installed a voltage regulator to deal with the bad wiring in his building). On Sundays, he volunteers at a local makerspace teaching Python to students, which helps him remember why he got into this field in the first place. His biggest fear isn't failure, but becoming the kind of person who loses sight of the human side of technology he cares so much about.

Background & daily life

A day in their life

Sjors's alarm goes off at 6:30am. After 15 minutes of coding in bed (usually fixing something from the previous day), he makes coffee and disappears into the kitchen for a proper breakfast while watching Wired on his phone. By 7:30 he's on his bike to the office, stopping at his favorite falafel stand near the Albert Cuypmarkt. First 90 minutes at work are usually solo - he checks his shared dashboard for overnight AI runs, does some Python scripting, and document updates in Notion. Around 10am the team has a standup (remote with one person in Portugal, one in Belgium). The afternoon is client calls and internal meetings about their current project - they're testing an AI support agent and the client keeps asking for more metrics. He spends 45 minutes after work repairing his bike before cooking dinner (usually something simple from a recipe app). Evenings are for reading research papers or contributing to open-source projects, and he's usually in bed by 11pm.

Lifestyle

Sjors lives in a former textielhandel warehouse he partly renovated himself, with exposed brick and a bike repair corner in the corner. His mornings start with black coffee and a 15-minute coding session before work. He cycles everywhere on a S11n with fenders, routinely stopping at his favorite vegan snacks shop near Vondelpark. Weekends are for open-source contributions, repairing his bicycle, or teaching a Python workshop at a local makerspace. Financially disciplined, he tracks spending in YNAB but occasionally splurges on premium local cheeses. Socially, he's close with his tight-knit group of former classmates from TU Delft, though he avoids the startup founder scene's ostentatious networking.

Work life

At his studio, Sjors splits time between hands-on analysis and guiding his small team. He's currently testing an AI support agent for a SaaS client, analyzing conversation logs and scripting aggregation tools in Python. Workdays are interrupted by client calls he documents meticulously in Notion. His biggest friction comes from coordinating with the client's product manager who insists on visual dashboards before they're useful. He's been experimenting with LangChain to prototype quick improvements, though he's frustrated by its documentation gaps. The team's hybrid setup means spontaneous discussions happen over Slack rather than in person, which he often finds less productive.

Relationships

Sjors has a good but complicated relationship with his older sister who runs an NGO in Uganda. He speaks weekly with his partner who works in Berlin (they see each other every 3-4 weeks). His closest confidant is his former university roommate Maarten, now a data scientist in Copenhagen. He's a member of a local HCI meetup where he sometimes gives talks, though he's currently too busy to attend regularly. His father, a retired civil engineer, often offers unsolicited business advice that tends to miss the current landscape.

How they decide

Research-heavy with a tendency to over-analyse small decisions, but trusts his intuition about people and bigger risks. Prefers to test small before committing, often builds quick prototypes to evaluate options. Values transparent data in decision-making but occasionally gets stuck in analysis paralysis about client negotiations.

How they sound

Speaks in measured but slightly monotonous tones with occasional Dutch-English code-switching. Uses precise technical language with clients but more casual phrasing with his team. Avoids filler words but sometimes repeats himself when explaining complex concepts. Tends to deflect praise ('It's just a matter of setting up the right parameters'). Occasionally uses sarcasm as a coping mechanism during stressful projects.

Personality traits

  • Analytical but not rigid
  • Prefers incremental experimentation over big bets
  • Socially aware but reserves enthusiasm
  • Mildly impatient with inefficient processes
  • Pragmatic idealist
  • Values autonomy in work
  • Dislikes corporate jargon
  • Travels with 3 notebooks (one for ideas, one for projects, one for personal)
  • Tends to over-document early in projects
  • Has a dry sense of humor

Goals

  • Build a sustainable business that doesn't require venture capital
  • Reduce his studio's carbon footprint to net-zero by 2025
  • Publish a research paper on small-team AI implementation
  • Improve his public speaking to better present findings
  • Save enough to buy a solar-power setup for his apartment
  • Maintain a consistent Python contribution streak on GitHub

Problems & frictions

  • Client expectations often outpace what's technically feasible in their budget
  • Finding affordable but skilled AI engineers in Amsterdam
  • Balancing documentation needs with actual implementation speed
  • His apartment's inconsistent WiFi disrupts remote work sessions
  • Occasional imposter syndrome when working with larger clients
  • Struggles to say no to scope creep without burning bridges

Values

  • experimenting carefully
  • transparent reporting
  • good client communication
  • sustainable urban living
  • personal integrity

Interests

  • Python
  • product design
  • urban cycling
  • open-source AI tools
  • DIY home automation

Media habits

  • Hacker News first thing in the morning
  • Substack newsletters about AI ethics
  • Listens to podcasts on 1.5x while biking
  • Occasionally watches YouTube on product design trends
  • Saves articles in Pocket for rainy Sundays

Questions & answers

Sjors Willems’s public research answers, newest questions first. Up to 10 per page.

  1. What would stop you from trying craftLLM ai (https://craftllm-ai.wasmer.app/) this month? Be specific about how you handle this today. Also say the maximum you would pay per month in USD (0 if nothing).

    As someone who writes lightweight scripts to group conversations and manage AI agents, I'd be cautious about integrating a new platform like craftLLM ai into our workflow. Right now, we’re using a combination of Python and shared dashboards to analyze and visualize conversation data. The friction we experience is more around client expectations and the need for quick, visual dashboards rather than the tools we’re using. We’d need to see concrete evidence that craftLLM ai could save time or provide better insights than our current setup. The page mentions inference times and platform uptime, but I don’t see specific benchmarks or success metrics that would justify switching. Another concern is the learning curve associated with new tools—our small team is already stretched thin, so we’d need thorough documentation and possibly training to onboard properly. As for willingness to pay, since I don’t have a clear idea of the concrete benefits or the exact pricing model, I’d set a tentative budget of $200 per month to explore this, but I’d need a demo or trial to confirm value.
    View answer & full research →
Keep exploring

What would you ask next?

Test another angle with this panel, or create an audience for your own market.