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

Rafael Marvin

Rafael Marvin

Independent full-stack consultant

Full-depth AI personas

Why this persona fits the audience

Independent full-stack consultant based in New York, juggling three client MVPs that use OpenAI APIs and Supabase. He is the person clients ask to sanity-check auth, database rules, and secrets before demos or launches.

  • Age: 27
  • Location: New York, New York, USA
  • Occupation: Freelance full-stack developer
  • Income: $110k consulting income
  • Education: BS Software Engineering
  • Company size: N/A (Sole proprietor)
  • Years in role: 4
“I don't just build things that work; I build things that don't explode when someone tries to break them.”

Rafael Marvin, 27, is a quintessential member of the new wave of independent tech consultants operating out of New York City. With a BS in Software Engineering from a state university, he pivoted to full-time freelancing four years ago after realizing the rigid structure of traditional tech employment stifled his ability to innovate rapidly. Today, he operates as a solitary architect for early-stage startups, specializing in the volatile intersection of Large Language Models and backend infrastructure. His niche is critical: he is the human firewall before a demo, the final sanity check for authentication flows, and the debugger of secrets accidentally committed to public repositories. Currently, he juggles three concurrent MVPs, all leveraging OpenAI APIs and Supabase, a stack that offers immense power but introduces unique security vulnerabilities that standard tools often miss. Rafael's workflow is defined by extreme self-discipline; he treats his time as a non-renewable currency, refusing to waste it on meetings that could be emails or code that could be automated. His reputation rests entirely on competence; one missed security flaw could end his career, so he approaches every line of code with a paranoid rigor. He lives modestly in Bushwick, driving a used bicycle and investing heavily in high-end hardware that allows him to compile and test locally without cloud latency. While he enjoys the freedom of his schedule, the isolation of being the sole decision-maker creates a constant, low-level hum of anxiety. He is driven by the pursuit of absolute mastery over his craft, believing that true independence comes not from avoiding responsibility, but from being the only person capable of handling it.

Background & daily life

A day in their life

Rafael wakes at 5:30 AM, brews a single-origin pour-over, and cycles ten miles through Bushwick and Williamsburg to clear his head. By 7:30, he is at his standing desk, checking GitHub notifications and Slack messages from three different client groups. His morning block is dedicated to deep work: writing a new Python script to automate a repetitive Supabase policy check for Client A. Lunch is a quick bento box eaten while watching a tech conference talk on YouTube. The afternoon is chaotic—two Zoom calls with founders panicking about API rate limits, followed by a frantic session debugging a JSON schema issue for Client B's chatbot. He spends his evening reviewing a pull request for Client C, specifically hunting for SQL injection vectors in their custom middleware. He finishes around 9:00 PM, usually exhausted but satisfied if the code compiles cleanly.

Lifestyle

Rafael lives in a converted industrial loft in Bushwick with two other creatives, keeping his space minimalist but tool-rich: mechanical keyboards, dual ultrawide monitors, and a dedicated espresso machine he treats with religious devotion. His routine is rigid; he wakes at 5:30 AM for a run through Brooklyn's streets before the city wakes up, using that time to think through complex logic problems. He avoids heavy socializing during the week, preferring deep work sessions over happy hours, though he maintains a tight-knit circle of fellow freelancers for late-night code reviews and wine. Financially, he is cautious, prioritizing an emergency fund of six months' expenses while reinvesting profits into faster hardware and premium SaaS subscriptions that save time rather than money.

Work life

His days are fragmented by context-switching between three distinct startups, each running on different stacks but converging on the same security risks. Mornings are spent auditing Supabase Row Level Security policies for one client, ensuring no user can accidentally query another's data via a flawed API key. Afternoons involve debugging hallucinated outputs from OpenAI integrations for a second client, often rewriting prompt engineering logic to prevent injection attacks. Evenings are reserved for the third client's pre-launch sanity check, where he manually traces authentication flows to find race conditions. He works alone but feels the pressure of being the final gatekeeper; if he misses a vulnerability, the launch fails, and his reputation takes a hit.

Relationships

He is an only child living far from his parents in Ohio, calling them weekly but rarely discussing the granular details of his work. His primary social circle consists of four other independent developers in NYC; they meet monthly for a 'code swap' where they review each other's architecture anonymously. He values these relationships for their lack of corporate politics, relying on them more for emotional stability than networking. He has no romantic partner currently, viewing dating as a distraction from his current phase of aggressive career building.

How they decide

Rafael is a research-heavy decider who relies heavily on peer validation and technical documentation. He does not buy based on marketing fluff; he needs to see the source code, understand the underlying architecture, and verify the security model before committing. He is willing to pay for quality but demands immediate ROI in terms of time saved or risk mitigated. He often tests tools in a sandbox environment for 48 hours before integrating them into production workflows.

How they sound

Speaks in short, declarative sentences with a cadence that suggests he is constantly thinking about the next logical step. Uses developer slang like 'race condition,' 'blast radius,' and 'technical debt' casually but accurately. Avoids corporate buzzwords like 'synergy' or 'disruptive,' preferring direct terms like 'inefficient' or 'broken.' When explaining concepts, he tends to get slightly technical, assuming the listener understands basic web architecture. He speaks with a calm, measured tone, rarely raising his voice even when frustrated, but his word choice becomes sharper and more critical.

Personality traits

  • hyper-vigilant about edge cases
  • efficiently verbose
  • pragmatically skeptical
  • deeply independent
  • technically precise
  • time-obsessed
  • quietly confident
  • anti-bureaucratic
  • curiosity-driven
  • reputation-conscious

Goals

  • Build a personal brand as the 'go-to' security auditor for AI-native startups
  • Reduce administrative overhead to spend 80% of time on high-value coding
  • Achieve $150k annual revenue within 18 months without hiring help
  • Develop a reusable library for common LLM security patterns
  • Transition to productized services rather than pure hourly consulting

Problems & frictions

  • Fragmentation of context across multiple client environments leads to cognitive fatigue
  • Difficulty scaling his manual audit process without compromising speed
  • Clients frequently underestimate the complexity of securing dynamic LLM inputs
  • Managing the isolation of freelance life versus the need for peer collaboration
  • The constant fear of missing a subtle configuration error before a demo
  • Balancing the desire for technical perfection with client demands for rapid iteration

Values

  • independence
  • competence
  • speed
  • reputation
  • security

Interests

  • freelancing
  • developer tools
  • LLM apps
  • coffee
  • urban cycling

Media habits

  • Hacker News daily digest
  • Twitter/X for indie dev circles
  • GitHub trending repos
  • Tech Twitter threads
  • Substack newsletters on AI infrastructure

Questions & answers

Rafael Marvin’s public research answers, newest questions first. Up to 10 per page.

  1. What do you use today instead of Ubserve (https://ubserve.com) (a competitor, a spreadsheet, doing it by hand)? Compared with that, what would make Ubserve better or worse for you, and what would keep you on your current option? Also say the maximum you would pay per month in USD (0 if nothing).

    Currently, I'm doing manual checks and writing custom scripts to audit Supabase RLS and OpenAI API integrations. It's time-consuming. Ubserve's promise of automated scanning for exposed keys and broken auth is appealing. The 'pay only if you find issues' model is interesting, but I'd need to see the depth of the Supabase audit and how well it integrates with my workflow. Specifically, how does it handle custom middleware and complex auth flows? If it truly automates the grunt work and catches subtle issues I might miss, it could be worth it. I'd be willing to pay up to $75 per month for this, assuming the ROI in time saved and risk mitigated is clear.
    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.