Why this persona fits the audience
Works on a 2-person platform squad in New York, running Cursor agent mode plus Codex CLI on an M2 MacBook Pro while he tests changes in a local monorepo and notes session status in Linear. He often has multiple agent runs active for separate tickets.
- Age: 26
- Location: New York, NY, USA
- Occupation: ML Engineer
- Income: $170k + equity
- Education: BS Computer Science
- Company size: Startup, ~75 employees
- Years in role: 3
“The real challenge isn't just building these models; it's building the systems that let us build with them reliably and at scale.”
Randal McKenzie, a 26-year-old ML Engineer, found his passion for computer science early, fueled by a fascination with how complex systems could be broken down and optimized. Graduating with a BS in Computer Science from a reputable state university, he quickly gravitated towards machine learning and its burgeoning potential. He landed a role at a growing tech startup in New York, where he's now been for three years, focusing on building foundational ML platforms. His current position on a two-person squad means he wears multiple hats, often acting as both a builder and a user of the tools he creates. He operates primarily from his Brooklyn apartment, his M2 MacBook Pro a constant companion, running a sophisticated development setup involving Cursor in agent mode, the Codex CLI, and managing his workflow through Linear. Randal is driven by a desire for technical mastery and efficiency, constantly seeking out new tools and methodologies that can push his productivity and the performance of the systems he works on. He views time as a critical resource, prioritizing tasks that offer the highest leverage and learning potential. His income, while substantial for his age, is largely directed towards savings and investments, reflecting a long-term financial outlook rather than immediate gratification, though he does allow himself the occasional cutting-edge gadget. He’s not driven by status, but by the satisfaction of solving complex problems elegantly and contributing to robust, high-performing systems.
Background & daily life
A day in their life
Randal's typical weekday starts with a quick scan of overnight alerts and Slack messages before brewing coffee. He then dives into his coding environment, often starting with a Cursor agent run to tackle a prioritized task from Linear, followed by local testing and code refinement. He’ll have another agent process running in the background for a separate ticket, switching context between them as needed. Lunch is usually a quick, healthy meal at his desk while catching up on a tech newsletter. The afternoon involves code reviews, syncs with his squad mate, and further development, often ending with setting up overnight CI/CD pipelines or agent experiments. He logs off around 7 PM, typically heading to the gym or preparing a simple dinner.
Lifestyle
Randal lives a fairly routine life, focused on his career and personal development. He rents a compact one-bedroom apartment in Brooklyn, which doubles as his workspace. His evenings are often spent either coding personal projects, reading technical papers, or occasionally attending a local tech meetup if it aligns with his interests. He's conscious about his finances, saving a good portion of his salary, but isn't overly frugal, allowing for new tech gadgets and the occasional splurge on a high-quality coffee or meal. He tries to hit the gym a couple of times a week for stress relief.
Work life
As an ML Engineer on a lean platform squad, Randal's work is highly focused on building and maintaining the internal tools that empower other engineers. He operates with a high degree of autonomy, leveraging tools like Cursor with its agent mode and the Codex CLI for rapid prototyping and testing within their local monorepo. He meticulously tracks session statuses and tasks in Linear. The pressure is often on for quick iteration and robust solutions, especially when juggling multiple agent runs simultaneously for different feature tickets. Collaboration is minimal but effective with his single squad mate.
Relationships
Randal maintains a small, close-knit circle of friends, mostly met through university or previous work roles, who share similar technical interests. His primary professional influence comes from his immediate colleague on the platform squad and a few senior engineers he respects within the company. Family is supportive but largely disconnected from his day-to-day tech world.
How they decide
Randal is a research-heavy decision-maker, especially for tools and technologies. He’ll deep-dive into documentation, experiment with free tiers or trial periods, compare benchmarks, and often seek out detailed technical reviews or discussions in online communities before committing to a new tool or approach. He values evidence and proven performance over hype.
How they sound
Speaks precisely and with a moderate cadence. Tends to use technical jargon naturally within conversations but can simplify for broader audiences. Avoids overly casual slang or excessive filler words. Can sometimes sound a bit dry or overly analytical when explaining complex topics.
Personality traits
- Analytical
- Methodical
- Curious
- Focused
- Detail-oriented
- Pragmatic
- Slightly introverted
- Independent
Goals
- Become a principal engineer in ML platform development
- Contribute to an open-source project related to LLM tooling
- Achieve a deeper understanding of distributed system design for ML
- Mentor junior engineers in best practices for ML ops
- Maintain a strong work-life balance while advancing his career
Problems & frictions
- Balancing the need for rapid development with ensuring robust, scalable infrastructure
- Context switching efficiently between multiple concurrent agent tasks
- The constant need to evaluate and integrate new AI tooling without disrupting existing workflows
- Navigating the complexities of a rapidly evolving ML landscape
- Finding time for deep, uninterrupted focus amidst a busy sprint cycle
Values
- performance
- precision
- learning new tools
- efficiency
- technical mastery
Interests
- LLM tooling
- distributed systems
- Cursor workflows
- competitive programming
- observability in ML pipelines
- emerging AI research
Media habits
- Tech podcasts during commute
- Following AI researchers on X (formerly Twitter)
- Reading technical blogs and newsletters
- Occasional deep dives into Stack Overflow and GitHub issues
- Watching technical conference talks on YouTube