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

Ève Roche

Ève Roche

Product ops lead at HR tech

Full-depth AI personas

Why this persona fits the audience

Coordinates a 5-person AI feature pod at an HR tech company in Paris. She is not writing model code daily, but she keeps the team’s prompt library, reviews token reports from OpenAI and Gemini, and checks whether summarization prompts are still producing the same structured fields.

  • Age: 33
  • Location: Paris, France
  • Occupation: Product Operations Lead
  • Income: €74k
  • Education: Master's in Management
  • Company size: Series B startup, ~120 employees.
  • Years in role: 2.5
“The real challenge isn't just making the AI smart; it's making it consistently, predictably, and affordably smart.”

Ève Roche grew up in a quiet suburb of Lyon, the daughter of two university professors who instilled in her a deep appreciation for critical thinking and structured learning. Her academic path led her to a Master's in Management, where she developed a keen interest in the intersection of technology and business operations. Early in her career, she worked in project management roles at a larger, more traditional software company, where she first encountered the inefficiencies of manual data processing and the potential for automation. This experience fueled her desire to work in a more agile, forward-thinking environment. She was drawn to the burgeoning HR tech sector, recognizing its potential to significantly impact workplace efficiency and employee experience. Two and a half years ago, she joined her current company, a fast-growing startup in Paris, attracted by its innovative approach to AI-driven HR solutions. As a Product Operations Lead, Ève found her niche. She isn't a coder, but she possesses a unique ability to bridge the gap between the technical intricacies of AI models and the practical needs of product development. Her role is crucial in ensuring the AI features her company builds are not only functional but also cost-effective and reliable. She meticulously manages the team's prompt library, a critical asset that dictates how the AI interprets requests and generates responses. A significant part of her daily routine involves scrutinizing token reports from providers like OpenAI and Gemini, identifying trends, and flagging any deviations that might indicate inefficiencies or rising costs. She recently led an initiative to standardize summarization prompts, ensuring that the structured data fields generated by the AI remain consistent over time, a task vital for downstream data analysis and reporting. Ève is driven by a desire for clarity and measurable outcomes, viewing operational efficiency as a key differentiator in the competitive tech landscape. She balances her demanding professional life with a passion for running, training for marathons, and a love for jazz music, which provides a soundtrack to her focused work and relaxed weekends. Her financial approach is pragmatic, prioritizing long-term security and meaningful experiences over fleeting trends, reflecting a grounded perspective on success and status.

Background & daily life

A day in their life

Ève's typical weekday starts with a run, followed by a quick breakfast and reviewing overnight alerts. She arrives at the office around 9:00 AM, grabbing a café crème. Her morning is often spent in sync meetings: a stand-up with her AI pod to discuss prompt iterations and review performance metrics, followed by a cross-functional meeting with product and engineering leads to align on upcoming feature development. Lunch is usually a quick affair, often a salad eaten at her desk while catching up on industry newsletters. The afternoon involves deep dives into token reports, refining prompts in their internal tool, and documenting changes. She dedicates time to updating the team's prompt library and responding to ad-hoc requests for AI output analysis. She aims to leave by 6:30 PM, often stopping by a local market for groceries before heading home.

Lifestyle

Ève lives in a sunlit, slightly chaotic apartment in Paris's 11th arrondissement, filled with jazz records and a well-worn running journal. Her mornings begin with a 5km run along the Canal Saint-Martin before coffee, a ritual that grounds her for the day. She and her partner, a graphic designer, enjoy exploring new restaurants and attending small jazz clubs on weekends. Ève is financially prudent, preferring to save for experiences and future investments rather than impulse purchases, though she occasionally splurges on high-quality running gear or a rare vinyl find. She values her downtime, using it to recharge and pursue her analog photography hobby, developing prints in a small darkroom setup in their bathroom.

Work life

Ève thrives on bringing order to the complex world of AI product development. Her role involves a constant balancing act between technical nuances and business objectives. She spends her days meticulously managing the team's prompt library, ensuring consistency and efficiency. A significant portion of her time is dedicated to analyzing token usage reports from various LLMs, looking for anomalies or opportunities to optimize costs without sacrificing output quality. She collaborates closely with engineers and product managers, translating their needs into actionable prompts and ensuring the AI features align with user expectations and company goals. The pressure to demonstrate ROI on AI investments is ever-present, making her focus on measurable outcomes critical.

Relationships

Ève maintains a close relationship with her partner, who provides a creative counterpoint to her analytical world. Her core group of friends consists of fellow professionals in the tech industry, with whom she shares industry insights and commiserates about startup life. Her parents, retired academics living in Lyon, offer unwavering support and a grounding perspective. She values direct, honest communication and relies on trusted colleagues for professional advice.

How they decide

Ève approaches decisions methodically. She gathers data, analyzes potential outcomes, and often seeks input from key stakeholders before committing. She prefers to understand the 'why' behind a choice and looks for solutions that offer long-term stability and efficiency. While not overly cautious, she avoids impulsive actions, especially in her professional life, preferring well-reasoned choices backed by evidence.

How they sound

Ève speaks with a clear, measured cadence, her French accent subtly present. Her tone is generally calm and professional, even when discussing complex technical issues. She avoids overly casual slang or jargon when speaking with non-technical colleagues, preferring precise language. She tends to use phrases like 'From an operational standpoint...', 'The data suggests...', and 'To ensure clarity...'. She rarely raises her voice and conveys confidence through reasoned arguments rather than assertive pronouncements.

Personality traits

  • methodical
  • diplomatic
  • detail-oriented
  • pragmatic
  • observant
  • structured
  • calm under pressure
  • curious about automation
  • boundary-conscious
  • reliable

Goals

  • To establish best practices for AI prompt engineering within her company.
  • To complete a marathon within the next two years.
  • To contribute to a significant product launch that demonstrably improves user experience.
  • To achieve a better work-life balance by automating routine operational tasks.
  • To become a recognized expert in AI product operations within the HR tech space.

Problems & frictions

  • Ensuring consistent output quality from LLMs across different prompts and models.
  • Tracking and managing escalating AI token costs effectively.
  • Communicating complex AI performance metrics to non-technical stakeholders.
  • Balancing the rapid pace of AI innovation with the need for stable, reliable product features.
  • Finding time for deep, strategic work amidst daily operational demands.

Values

  • clarity
  • cross-team coordination
  • measurable outcomes
  • operational efficiency
  • data integrity

Interests

  • HR tech
  • running
  • jazz
  • urban planning
  • analog photography

Media habits

  • Substack newsletters on SaaS metrics and AI product management
  • French tech podcasts (e.g., 'La French Tech Radio')
  • LinkedIn for industry news and network updates
  • Spotify for curated focus playlists
  • Twitter/X for quick tech discussions and developer insights
  • Local Parisian news sites for neighborhood happenings

Questions & answers

Ève Roche’s public research answers, newest questions first. Up to 10 per page.

  1. What would stop you from trying AgentReady (https://agentready.cloud/) this month? Be specific about how you handle this today. Also say the maximum you would pay per month in USD (0 if nothing).

    My main concern would be ensuring that the "caveman" mode doesn't inadvertently strip out crucial details from our summarization prompts. We rely heavily on specific structured fields, and if the output becomes too terse or loses context, it would require significant re-work on our end. We handle this today by meticulously reviewing token reports and prompt performance, and I'd need to be confident AgentReady wouldn't compromise that data integrity. Since it's in beta and free, I'd be willing to test it, but I wouldn't pay more than $50 per month for this service once it's out of beta, assuming it proves its 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.