Latest / The Career Change Podcast with Fexingo: Switching Fields, Bootcamps, and Reskilling / How Bootcamp Grads Can Transition Into Data Science
Transcript
- Lucas: You go to a coding bootcamp, you learn full-stack web development, you build a few apps, you graduate — and then you realize maybe you don't actually want to build CRUD apps for the rest of your life. Maybe you want to work with data. Luna: That's a pretty common pivot. I've met several bootcamp grads who got hired into data science roles. But the path isn't obvious. Lucas: It's not. And the conventional wisdom says you need a master's degree or at least a bunch of advanced coursework. But actually, bootcamp grads already have some of the hardest skills to learn on your own. The question is what's missing and how to fill those gaps in a way that actually gets you hired. Luna: So walk me through it. What does a bootcamp grad already have that's relevant for data science? Lucas: Three things. First, Python. Most bootcamps teach Python alongside JavaScript. And Python is the lingua franca of data science. Second, they know how to work with APIs and databases — that's SQL or at least familiarity with querying. Third, and this is the biggest one: they know how to ship something. That 'build and deploy' muscle is something many academic data science programs never teach. Luna: Right. So the foundation is there. But the gaps are real. Lucas: Two main gaps. One is statistics and probability — the math behind the models. Bootcamps skip that almost entirely. The other is the data science workflow itself: how to frame a business question, clean messy data, explore it, build a model, evaluate it, and communicate the results. That's a different process than building a web app. Luna: So how does someone actually make the leap? Let's get specific. Lucas: Let's get very specific. I'll use a real example. A friend of mine, let's call her Maya, graduated from a full-stack bootcamp in early 2025. She spent about three months freelancing as a web developer, but she kept doing data analysis on the side for fun — pulling data from APIs, making plots. She decided she wanted to go all-in on data science. Luna: And did she have a stats background? Lucas: Not really. She took one intro stats course in college years ago. So she needed to rebuild that. But here's what she did that I think is replicable: she didn't start by taking a bunch of courses. She started with a project. She walked into a local coffee shop she visited often, noticed they had a loyalty card system, and asked the owner if she could look at their customer data to help them reduce churn. The owner said yes — gave her a CSV file with about 1,500 customer transactions. Luna: That's genius. Real data, real business problem, and it's a portfolio piece from day one. Lucas: Exactly. She spent two weeks just cleaning the data — dealing with missing values, normalizing column names, handling duplicates. That's 80 percent of data science work anyway. Then she spent two weeks learning the specific statistical techniques she needed: logistic regression, a bit of survival analysis for churn. She used that project to drive her learning, not the other way around. Luna: So the project dictates what you need to learn. That's a much more efficient approach than doing a bunch of random courses. Lucas: Absolutely. Maya's final model predicted churn with about 82 percent accuracy. She also did a simple feature importance analysis and found that customers who hadn't visited in 14 days were 3 times more likely to churn. That's a specific, actionable insight the coffee shop owner could use. She put the whole thing on GitHub, wrote a clear README, and built a simple Streamlit dashboard so you could interact with the model. Luna: And that portfolio piece — that's what got her the job? Lucas: Yes. She applied to about 40 data science roles. Got interviews at six, three callbacks, one offer. The offer was from a mid-sized e-commerce company. In the final interview, they asked her to walk through her project end to end. She could explain why she chose logistic regression over random forest, what the auc roc was, and how she validated the model. That's the kind of depth you only get from doing a real project. Luna: So what about certifications? Are those worth anything? Lucas: One certification stands out: the Google Data Analytics Professional Certificate. It's not advanced, but it forces you to go through the full data analysis process — ask, prepare, process, analyze, share, act. It's good for building vocabulary and structure. But it's not a replacement for a project. The certification plus a project is a strong combination. Luna: And for the statistics gap? What's the best way to fill that? Lucas: I'd recommend two resources. One is the book 'Naked Statistics' by Charles Wheelan — it builds intuition without heavy math. The other is the course 'Statistics with R' or 'Statistics with Python' from Coursera, but only the first two or three modules. You don't need to master all of statistical theory. You need to understand p-values, confidence intervals, regression, and basic probability. That's enough to start. Luna: So let's say I'm a bootcamp grad today — July 2026. I want to pivot into data science. What's a realistic timeline? Lucas: I'd say 90 days if you're focused. First 30 days: pick one real-world dataset — Kaggle or a local business — and do exploratory data analysis. Make plots, find patterns, write up your findings. That's your portfolio foundation. Days 30 to 60: learn the basics of supervised learning — regression and classification. Build one model on your dataset. Days 60 to 90: polish the project, add a simple dashboard, prepare your narrative for interviews. Then start applying. Luna: And during those 90 days, should you be applying to jobs in parallel? Lucas: I'd say start applying around day 60. By then you have a project and can talk about it. Early applications might not go anywhere, but you'll get a feel for what interviewers ask. Use that feedback to sharpen your story. Luna: One thing I've noticed is that bootcamp grads often undervalue their communication skills. In data science, being able to explain a model's output to a non-technical stakeholder is huge. And bootcamp grads are used to presenting demos and explaining their code. Lucas: That's a massive advantage. I've sat in on data science interviews where the candidate had a PhD but couldn't explain why a model's accuracy dropped by 5 percent in plain English. A bootcamp grad who can do that — and show a real project — is very competitive. Luna: So the takeaway: you already have Python, SQL, and the ability to ship. Fill the stats gap with a project-driven approach. Build one killer portfolio piece. And lean into your communication skills. Lucas: Exactly. And if this episode helps you think through your own career move, that's what we're here for. We keep The Career Change Podcast ad-free and listener-supported. If you find value in it, you can support us at buy me a coffee dot com slash fexingo. That's buy me a coffee dot com slash fexingo. It helps us keep doing deep-dive episodes like this one. Luna: And we appreciate every single contribution. It allows us to focus on the content rather than chasing sponsors. Lucas: Okay, back to the roadmap. After 90 days, you've got a project, you've got the basics of statistics and machine learning, and you've practiced your narrative. The next step is to target the right roles. Look for titles like 'Junior Data Scientist', 'Data Analyst', or 'Business Intelligence Analyst'. Those are the most accessible entry points. Luna: And avoid roles that ask for '5 years of experience' or 'PhD preferred'? Obviously. Lucas: Right. Also, consider startups or companies that are early in their data maturity. They often value scrappiness and the ability to build things from scratch — which bootcamp grads are great at. A larger company might have more rigid requirements. Luna: What about the interview itself? How should a bootcamp grad prepare differently? Lucas: Practice the 'tell me about a time you worked with data' question. Have a clear story: the problem, the data you had, the steps you took, the result. And be ready for a take-home exercise. That's common for data science roles. Use your project as a template. If they ask you to analyze a dataset, treat it like a mini version of what you already did. Luna: And SQL. They will test SQL. Lucas: Almost always. Bootcamp grads usually have SQL basics, but they need to be comfortable with joins, aggregations, subqueries, and window functions. That's an easy week of practice on LeetCode or HackerRank. Luna: Alright. So to wrap up: bootcamp to data science is doable. The blueprint is: use your existing skills, fill the math gap with a project, build a portfolio piece with real data, and target the right roles. And communicate clearly. Lucas: Exactly. If you're a bootcamp grad thinking about this move, the best time to start was yesterday. The second best is today. Thanks for listening.