How I Built a CustomGPT to Represent My Non-Linear Career
On uploading my professional accomplishments to ChatGPT
We’ve entered the era where non-linear careers can be trained into AI models.
Not in the “upload your consciousness to the cloud” kind of way (yet?) but in something more practical: designing a personal AI that reflects how you’ve thought, worked, and solved problems throughout your career.
Mine is a customGPT of my portfolio career trajectory.
Why I Built It
For most of us, a resume or LinkedIn profile is the official record of our careers. However, both of these flatten the story. Each is a static document that strips away the nuance, the lived experiences, the strategy behind each move.
I’ve been an operator, strategist, chief of staff, political theorist, historian, and startup mentor. I’ve advised billionaires and integrated workforces during mergers. I’ve led transformations in hedge funds and industrial giants, while also mentoring founders through Oxford Entrepreneurs Network and CamEntrepreneurs.
A static resume can’t hold all that complexity, because I don’t end up landing the role.
I’m both overqualified, and somehow don’t have enough directly relevant experience…
What It Represents
I trained a customGPT on the various threads of my career:
Transformation projects: HR M&A, building out a holding company, workforce strategy at a startup, payroll and benefits harmonizations in a legacy organization, audit remediation, multimillion dollar cost savings initiatives.
Operator instincts: fast, intelligent, and disciplined execution, all in very high-pressure environments.
Intellectual foundation: Degrees from Columbia and Cambridge, in political thought and intellectual history, through the lens of power and reinvention, not to mention my major academic publications.
Venture Capital Diligence: I even used ChatGPT’s Deep Research feature to forensically reconstruct my VC experience from publicly available information on the internet, and wrote this essay about that professional chapter.
Entrepreneurial edge: What I’ve learned running my own LLC, mentoring startups and founders, building a portfolio career in real time…
Instead of telling people what I’ve done, I can let the model speak with my AI voice: advising on workforce integration, managing a CEO transition from the Austrian alps, or even writing about luxury plane linens as metaphors for class and control.
The Process
And building a customGPT doesn’t require any code; all that’s needed is a good prompt, and some concrete data-rich evidence. I fed it:
Links to my essays and newsletters (the real-time thinking)
Career documents and strategy memos (the frameworks)
Reflections, case studies, stories, and my Myers-Briggs assessment results (the human side)
The resulting interface is an interactive representation of my portfolio career, a living record can grow and develop as I continue in my professional journey.
What I Learned
A career is data-rich: Every deal, transformation, or failure contains a pattern. Training a model forces you to see those patterns clearly.
I credit two of my managers who were both ex-General Electric with teaching me how to quantify my impact at work. They explained that when performance review time came at GE, the person reading your evaluation had no idea who you were… so the write up had better be succint and filled with demonstrated impact.
Voice matters. The most technical resume bullet points only came alive when I paired them with the narrative that sometimes can’t even be conveyed in an interview.
How you you adequately describe managing a CEO transition remotely from 4,000 miles away, if not with an extended narrative? Or the time I competed in a Triathlon as part of a team-building activity?
Operators > Influencers. I built this tool initially for hiring managers and recruiters, to match my skills and experiences with their job descriptions more easily. But in the end, I use it much more frequently than anyone else (or so I think?)
Why It Matters
In a world of layoffs, reorgs, and portfolio careers, touting your professional accomplishments isn’t optional. It’s table stakes when starting from scratch every 5 years, as anyone who’s been laid off will tell you…
A customGPT cannot replace me, but it does makes my experiences more scalable than 4 rounds of interviews. It makes my career legible to others while reminding me that the through-line is not any single role, but the system of thinking through the various mental models I’ve built across all of my roles and experiences.
I’ve come to believe that using AI to represent your accomplishments, especially the messy, nonlinear arc that predominates the careers of so many Millenials and Gen Z, is an undeniable edge. Give it a try, see what you think. Is uploading your professional accomplishments to ChatGPT a bridge too far? Reply to this email: I’d love to hear your thoughts.
👉 If you’re curious, subscribe along. This Substack is where I’m documenting not just what I’ve done, but how to make sense of your own portfolio career in a system that wasn’t designed for you.
And for all the paid subscribers to ARK Strategy, I break down the bespoke prompt I used to train the CustomGPT model below: