In March 2020, I couldn't write a single line of Python.
By 2023, I was building dashboards and automating reporting for a Google project, saving 10 hours a week of manual work.
By 2025, I was running my own Data & AI consulting practice from Tbilisi, Georgia.
This isn't a "learn to code in 30 days" story. It's messy, non-linear, and full of detours. But if you're thinking about pivoting into data or going independent, some of this might be useful.
Table of Contents
- The accidental start
- What the diploma didn't give me
- Getting into Google (through the back door)
- What I actually built there
- Why I left a comfortable position
- The consulting stack I use today
The accidental start
I grew up in the French Alps. Studied mechanics. Worked random jobs. No tech background whatsoever.
Then COVID hit, and like millions of people, I was stuck at home. I stumbled onto free data science livestreams by a French educator and something clicked immediately. Not in a "this seems like a good career move" way, but in a "this is exactly how my brain works" way.
I'd always been the kid who bypassed the family computer password through Safe Mode to play video games, or ran two separate school report booklets, one for teachers, one for parents, to game the system. That wasn't rebellion. It was systems thinking, at a teenager's scale.
Data turned out to be the same thing: find the structure, spot the gap, make it work. I enrolled in a Data Analyst program (RNCP Level 6, the French equivalent of a Bachelor's), learned Python, SQL, and data visualization, and graduated in 2022.
What the diploma didn't give me
A job.
In France, a diploma without experience is background noise. I applied everywhere. Nothing. The classic chicken-and-egg problem: you need experience to get hired, but nobody will hire you without experience.
So I did what I always do when the front door is locked. I went around.
I moved to Portugal. Not as a career strategy. As a reposition. Lower cost of living, different job market, and a bet that proximity to international companies would open something up.
It did.
Getting into Google (through the back door)
I landed a Business Analyst role at Teleperformance, assigned to a Google project. Not Google directly, a partner operation. But the standards, the culture, the data rigor: all Google.
The project was new internally. No established processes, no structured reporting, no dashboards. Basically a blank canvas with messy data.
That's my favorite kind of problem.
What I actually built there
Here's the concrete stuff, because dev.to isn't the place for vague corporate talk:
Automated reporting that saved 10h/week. The team was manually pulling numbers from multiple sources every week. I built SQL queries that aggregated everything into a single Looker Studio dashboard with scheduled refreshes. Nothing revolutionary. But for a team that was navigating blind, it was a paradigm shift.
Built cross-functional reporting. Operations, sales, and management were all looking at different numbers. I created a single source of truth. Sounds basic, but alignment on metrics is one of the highest-ROI things you can do in any organization.
Why I left a comfortable position
Google taught me rigor and scale. But I kept bumping into a ceiling, not of competence, but of scope. I wanted to choose which problems to solve, which companies to work with, and how deep to go.
In parallel, I'd moved to Tbilisi, Georgia. What was supposed to be a one-month trip turned into a life. The city has this raw energy, everything feels like it's being built from scratch. I met my partner Darina there, and some decisions just don't need a pros-and-cons list.
So I made the jump. Independent Data & AI consultant.
The consulting stack I use today
After a year of client work, here's what I actually use daily. Not what looks good on a resume, but what gets deployed:
For data strategy & dashboards:
- BigQuery for anything beyond basic analytics
- Looker Studio for client-facing dashboards (free, good enough for 90% of SME needs)
- Python + pandas for data cleaning and transformation
- SQL everywhere, always
For AI integration:
- Claude API and OpenAI API depending on the use case
- LangChain for RAG pipelines when clients need AI on their own documents
- Custom prompting architectures (the real value isn't the model, it's the system design around it)
For automation:
- n8n (self-hosted) for workflow orchestration
- Google Apps Script for anything living in the Google ecosystem
- REST APIs as the glue between everything
The tech isn't exotic. The value is in knowing which tool fits which problem, and building systems that compound over time.
I'm now based in Tbilisi, working with French and international companies on data strategy, AI integration, and automation. If you're on a similar path, whether pivoting into data, considering going independent, or just trying to figure out what stack to learn, feel free to reach out.
And if you're a company sitting on data you're not using, or manual processes eating your team's time: that's literally what I do.
What was your pivot moment? Drop it in the comments, I'm curious how other people ended up in tech.
Top comments (0)