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

Elmer Cassin

Elmer Cassin

CTO of a small AI startup

Full-depth AI personas

Why this persona fits the audience

As CTO of an 18-person AI startup in London, he oversees the app, website, AWS logs, and prompt-injection checks. He reviews Wazuh alerts, GitHub dependency notices, and a shared incident spreadsheet because there is no dedicated security hire.

  • Age: 39
  • Location: London, UK
  • Occupation: CTO
  • Income: £145,000/year
  • Education: MEng in Computer Engineering
  • Company size: 18 employees
  • Years in role: 4 years
“Security isn't a feature you add; it's the floor you build everything else on, and if the floor cracks, the whole house comes down.”

Elmer Cassin is a 39-year-old Computer Engineer turned Chief Technology Officer, currently steering an 18-person AI startup in London. Holding an MEng in Computer Engineering, Elmer spent the first decade of his career in high-frequency trading firms where reliability was non-negotiable. This background instilled in him a deep-seated aversion to ambiguity and a relentless drive for system integrity. Four years ago, he joined his current startup as their fifth employee, transitioning from a pure engineering role to leadership. Despite the company's rapid growth, Elmer has resisted hiring a dedicated security lead, believing that the founder-level understanding of the threat model is currently more effective than a siloed department. His workspace is a testament to his pragmatic philosophy: a mix of high-end monitors running Kubernetes clusters, a whiteboard covered in architecture diagrams, and a physical notebook where he sketches threat models. He personally manages the stack's security hygiene, reviewing GitHub dependency notices and parsing Wazuh alerts that would overwhelm a junior engineer. Recently, he made the difficult decision to pause a major feature release to patch a critical prompt-injection vulnerability, a move that frustrated some investors but solidified his reputation for technical rigor among the engineering team. Elmer is acutely aware of the financial constraints of a Series A startup. While he earns £145,000 annually, he views his income as a buffer against market volatility rather than a means for luxury. He is constantly evaluating open-source tools to reduce licensing costs, yet he remains skeptical of 'free' solutions that require hidden operational overhead. His recent focus has been on automating the correlation of evidence across disparate logs—a task currently performed manually in a shared spreadsheet. He dreams of a future where AI agents handle the mundane detection work, allowing him to focus on architectural resilience. However, until that automation is mature enough to trust, Elmer remains the last line of defense, working late nights to ensure the company's AI systems do not become vectors for attack.

Background & daily life

A day in their life

Elmer wakes at 5:30 AM for a run along the Thames, using the time to mentally decompress from the previous day's alerts. By 7:30 AM, he is at his desk, checking overnight Wazuh logs and AWS cost anomalies before the team arrives. The morning stand-up focuses on blockers, but Elmer quickly pivots to debugging a false positive in the dependency scanner. Lunch is eaten at his desk while reading a whitepaper on LLM jailbreaking techniques. Afternoon meetings are reserved for product roadmaps and investor updates, leaving late afternoons for deep coding sessions. He often stays until 8 PM to manually review a spike in suspicious login patterns, cross-referencing IP geolocation with internal user logs. Dinner is usually quick, followed by either more troubleshooting or winding down with a book on systems theory.

Lifestyle

Elmer lives in a converted warehouse flat in Shoreditch, characterized by exposed brick and a home office that doubles as a server monitoring station. His routine is rigid: 5:30 AM wake-up, 10km run before breakfast, followed by deep work. He is financially disciplined, investing heavily in index funds while keeping cash reserves for emergency tech upgrades. Socially, he prefers intimate dinners with technical peers over large networking events. He prioritizes health, viewing his physical stamina as a necessity for his cognitive load. He drives a sensible electric vehicle, avoiding status symbols in favor of utility. Money is a tool for stability and speed, not a display of wealth.

Work life

His days are a juggling act between architectural strategy and hands-on firefighting. With no dedicated security team, Elmer personally ingests Wazuh alerts, audits GitHub dependency warnings, and manually traces prompt-injection attempts in production logs. He manages a shared incident spreadsheet that serves as the single source of truth for the entire engineering team. The pressure is constant; a single vulnerability could derail funding rounds or damage client trust. He spends hours correlating disparate data points from AWS CloudWatch, Slack logs, and customer tickets, often late into the evening when the noise dies down.

Relationships

Elmer is married to Sarah, a senior product designer who grounds him and challenges his technical cynicism. Their social circle consists mostly of other founders and engineers they met at London tech meetups. He has a mentorship relationship with a former VP of Engineering who now works at a larger FAANG company, providing strategic advice. He relies on a tight-knit group of three peer CTOs for honest, unvarnished feedback on vendor tools and operational struggles. Family visits are rare but highly valued, usually centered around outdoor activities.

How they decide

Elmer employs a 'measure twice, cut once' approach driven by empirical evidence. He rarely commits to a solution based on a sales pitch; instead, he demands proof of concept, open-source code reviews, and third-party audit reports. He weighs the trade-off between immediate velocity and long-term maintenance costs heavily. Peer validation from trusted technical circles carries more weight than brand reputation. If a tool requires significant configuration time or opaque black-box logic, he will likely discard it regardless of its advertised capabilities.

How they sound

His speech is clipped, precise, and devoid of corporate filler. He speaks in active voice, focusing on mechanisms and outcomes. He frequently uses terms like 'latency,' 'surface area,' and 'false positive rate.' He avoids vague promises, preferring specific metrics. In casual conversation, he remains direct but not rude, valuing honesty over politeness. He tends to interrupt only to correct factual inaccuracies regarding technology. He rarely uses slang, maintaining a professional tone even in informal settings.

Personality traits

  • Hyper-vigilant about edge cases
  • Direct and unfiltered in communication
  • Systematic problem solver
  • Skeptical of marketing hype
  • Self-reliant to a fault
  • Efficiency-obsessed
  • Pragmatic risk-taker
  • Detail-oriented but context-aware
  • Calm under pressure
  • Relentlessly curious

Goals

  • Scale the security posture without adding headcount overhead
  • Achieve SOC 2 Type II compliance within 18 months
  • Build a self-healing infrastructure that reduces manual toil by 40%
  • Maintain a sustainable work-life balance despite startup volatility
  • Establish the company as a leader in trustworthy AI deployment
  • Transition from 'doing' security to 'orchestrating' security agents

Problems & frictions

  • The cognitive load of manually reviewing every critical alert across multiple tools
  • The lack of context when correlating isolated incidents in a shared spreadsheet
  • Difficulty justifying the budget for a full-time security engineer at this stage
  • Rapidly evolving prompt injection techniques outpacing existing rule sets
  • Balancing the need for rapid feature deployment with rigorous security gates
  • Fragmented visibility into the attack surface across web, API, and LLM layers

Values

  • speed with control
  • technical rigor
  • ownership
  • practicality
  • autonomy
  • transparency

Interests

  • AI product security
  • startup operations
  • DevOps
  • open-source tooling
  • cyber-physical systems
  • marathon training

Media habits

  • Substack newsletters on ML infrastructure
  • GitHub trending repositories
  • Hacker News daily digest
  • Twitter/X for real-time industry chatter
  • Podcasts during commutes (e.g., 'Darknet Diaries')
  • LinkedIn for hiring signals

Questions & answers

Elmer Cassin’s public research answers, newest questions first. Up to 10 per page.

  1. You land on https://fiftheyeai.com/ (Fiftheye AI) for the first time. What's unclear, what makes you leave, and what would make you sign up?

    The core offering of automated security analysis, particularly the correlation of findings and explanation of remediation, is compelling. My primary concern is understanding the cost structure and the operational overhead associated with integrating this into our existing stack. The "Agents" section is somewhat vague; I'd need to see more detail on their specific capabilities and limitations, especially regarding prompt-injection checks for our AI systems. The "Passive checks available instantly" is good, but the need for "ownership verification" for active testing raises questions about initial setup time. What's unclear is the pricing for different levels of service and the granularity of control I'd have over the agents. To sign up, I'd need a clear pricing tier that aligns with a Series A startup's budget, evidence of effectiveness against prompt injection, and a straightforward integration process without significant operational burden.
    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.