Don’t Waste 2026 on the Wrong Career - How to Pick the PERFECT Tech Role

Thanks to HubSpot for sponsoring this video. If you choose the wrong tech career path in 2026, you could waste years learning skills for a career you hate. But it can feel impossible to figure out what the right tech role is for you in 2026. Should you go into software engineering, data science, or maybe do something with AI? There are lots of different titles for roles that all code and kind of seem to be the same thing, but not. Today we're clearing up that confusion and giving you a simple framework to decide what path to pursue. I've worked across most of these roles myself, and I've mentored nearly 200 people into tech, so I have a pretty clear understanding of the right questions to ask yourself so you don't waste time studying the wrong thing. I'm going to walk you through a series of questions that will narrow down the possibilities to find the job that fits your strengths, preferences, and personality. Along the way, I'll break down each role in detail so you can actually understand what you're signing up for. Let's get started. The first question is the most important one, but luckily it's also the one most people can answer right away. These are two fundamentally different ways of spending your day. Builders create things. They write code that becomes a feature, a pipeline, or an application. At the end of the day, something exists that didn't exist before, and it's more concrete. Discoverers investigate things. They dig through data, test hypotheses, and find patterns. At the end of the day, there's an insight or a model that helps someone make a better decision. If you said building, we're going left on the tree. If you said discovering, we're going right. Let's start with the builders. The next question is about your relationship with ambiguity and change. Some builders love clarity. They want to write code, run it, and know for sure whether it worked. They value deterministic systems. Basically, if you use the same input, you get the same output every time. And they appreciate a structured career path where you know what senior looks like and how to get there. Other builders are comfortable or maybe even energized by uncertainty. They're okay working on problems where the output isn't predictable, where you might do everything right and the system still doesn't behave the way you expect, and where the tools and best practices are changing so fast that what you learned six months ago might already be outdated. They find that kind of challenge interesting rather than frustrating. This is a real personality difference. Neither is better, but it really changes which roles will feel rewarding versus draining. If you value deterministic systems and structured work, the next question is simple. Do you want to build products that users interact with or the infrastructure that everything else depends on? Some people are really motivated by seeing users engage with what they've built. So software engineering might be a really good fit. Others get more satisfaction from building the invisible foundation that makes everyone else's work possible. So they might be better as data engineers. Let's talk about software engineering first. Software engineers build pretty much everything you interact with digitally. The apps on your phone, the websites you browse, and the backend systems that keep everything running. Let me walk you through what this actually looks like. Say you're a software engineer at Spotify working on a new podcast sharing feature. Your day might involve starting with a product spec and architecture discussion, breaking down technical requirements like which APIs to use and cross-platform compatibility, writing code and tests while thinking about edge cases, and deploying changes while monitoring for errors. The skills you need include strong coding fundamentals in languages like Python, JavaScript, TypeScript, or Go, knowledge of data structures, algorithms, and system design, and experience with cloud services, containers, CICD, and monitoring. What makes this role great is exactly what we talked about, the clarity. Your code either works or it doesn't. You build things that actually exist and people use. It's the most well-defined of all these roles with clear specializations. Frontend, backend, full stack, mobile, DevOps, infrastructure, and a fairly straightforward career ladder. It's also one of the most accessible paths in. While degrees definitely help, some software engineers do start through bootcamps or self-study. You don't need advanced degrees to get your foot in the door. But I have to be honest about the current market. Entry-level positions are harder to get into than during the COVID boom, and with AI coding tools improving, employers are looking for everyone to have more senior architecture and system design skills. That being said, the BLS still projects 15% growth through 2034, so the long-term outlook is strong. One thing that all of the jobs on this list have in common is that they will be using AI tools regularly, whether that's for coding, writing, research, or whatever. And a lot of people are either under-utilizing these tools or using them inefficiently because they never learned how to prompt well. I suggest checking out this free guide called AI Decoded. It's a pocket guide to AI models, prompting frameworks, and useful GPTs. What I like about it is that it's practical and concise. It covers which AI models are actually good at what, because not every model is ideal for every task, and it has this nice prompting framework that structures your requests for better outputs. There's also a section on advanced prompting techniques and non-obvious tools you can use to make your workflows more efficient. My favorite part is probably the model comparison chart. When you're choosing between GPT, CLAW, Gemini, or open-source options like LLAMA, it helps to know their actual strengths rather than just defaulting to whatever you're used to. The guide is free, so if you want to level up how you're using AI day-to-day, I'll link that in the description. Now let's take a look at a role with a strong outlook even in the AI era. Data engineering. Data engineers build the systems that capture, clean, store, and deliver data across an organization. Every insight an analyst finds, every model a data scientist builds, every recommendation system a machine learning engineer deploys, all of it depends on data engineers having done their job well. Here's what that actually looks like. Imagine Netflix. Every second, millions of users are streaming, rating content, and searching. Data engineers are the ones who build systems to ingest and process millions of real-time events per second, design data models that track behavior across shared accounts, ensure data is properly indexed for fast access, and set up monitoring to catch quality issues before users are affected. The toolkit includes Python and SQL, distributed systems like Spark and Kafka, orchestration with Airflow, and cloud platforms. Like software engineering, the core goal is reliability and correctness. A pipeline either delivers the right data, or it doesn't. One of the realities of data engineering is that their best work goes completely unnoticed. Nobody comments when the pipeline works perfectly, so you may feel like you're always in a supporting role. But that might be worth it, because data engineering is probably the most AI-proof role on this entire list. Every AI system requires clean, well-structured data. McKinsey specifically names data engineers among the most in-demand AI-related hires, and it typically doesn't require an advanced degree. Now, if you're thinking you actually want to be a bit more in the thick of things with AI and all the changes right now, let's go back up the tree to the other side of the ambiguity split. If you're a builder who's comfortable with ambiguity, you've got two paths, machine learning engineering and AI engineering. Both involve working with AI systems that are inherently unpredictable. The question that separates them comes down to math and education. Do you enjoy calculus, linear algebra, probability, and statistics? And are you willing to invest in extensive education, typically a master's or PhD, to build that foundation? If yes, let's talk about machine learning engineering. Machine learning engineers take models and make them work reliably at scale in production. Here's an example of their day-to-day. Imagine you're an MLE at Uber and the data science team has built a surge pricing model. Now it's your job to make it return predictions in milliseconds, ensure it works 24-7 without crashing, handle millions of requests during peak hours, update the model without disrupting service, and monitor everything so nothing unexpected happens. The skills required are a unique combination. You need strong software engineering fundamentals, deep experience with machine learning, and system design for machine learning systems. The demand for this role is really high. Machine learning engineer postings grew 40% year-over-year in 2025, on top of 78% in 2024, making it one of the fastest-growing job titles in a major analysis of 180 million postings. But there are definitely challenges. The entry barrier is high. You realistically will need a master's or PhD, some related prior experience, and a strong network or portfolio. The field also moves incredibly fast, so you will never be done learning. And here's the thing a lot of people really don't understand before entering. Often you can do everything by the book, and the model still won't perform how you want. That's the ambiguity we talked about, and you have to be comfortable with it. If that level of math and education sounds like a lot, or if you just want to get into the AI space faster, let me tell you about AI engineering. AI engineering is the newest role on this list, and it is growing incredibly fast. The main difference between an AI engineer and a machine learning engineer is that AI engineers don't train models from scratch. Instead, they build applications on top of foundation models like Cloud or DeepSeek. Here's what a day might look like. Your company needs a customer service chatbot. You pick a foundation model, get a basic chat working via the API, write system instructions so the bot speaks in your company tone, connect it to real data through a RAG database, which means breaking up help center docs into chunks, embedding them, and letting the bot pull relevant information for each question. You build evaluation frameworks, add guardrails so it doesn't leak private data, and optimize for speed and cost. Now, people often say AI engineers are just software engineers calling a different API. And honestly, if you're a bad AI engineer, that's true. But a good AI engineer is a distinct and valuable role. They understand the landscape of foundation models and can evaluate which one is right for a given use case. They understand embeddings, machine learning evaluation metrics, and ranking algorithms. They know how to run systematic experiments, not just spot check outputs. And critically, they're comfortable building systems where the output is non-deterministic. LLMs don't always return what you expect, and you need to design around that. That's a big additional skill set on top of software engineering. AI engineer ranked number one on LinkedIn's jobs on the rise for two consecutive years. The on-ramp is faster than machine learning engineering because fewer jobs require a master's or PhD, and the expectations for foundational math and machine learning theory are lighter. Basically, most machine learning engineers can do AI engineering, but not all AI engineers can do machine learning without more depth and foundations. That means AI engineering is a faster path to get started. That covers the entire build branch of our tree. Now let's go back up the top and head right to the discover branch. If you said discover, your energy comes from analyzing data and uncovering insights. The next question is about how you like to structure your day. Do you prefer variety, responding to stakeholder needs, and having clear questions to answer? Or do you prefer long focus blocks, self-directed exploration, and defining your own problems? This is a real personality difference. Some people thrive on the fast pace of reactive work. You get new questions every day, direct stakeholder interaction, and clear deliverables. Others find that draining and want uninterrupted time to go deep on harder problems. Let's start with the more reactive, stakeholder-focused path, data analyst. Data analysts are like business detectives. When the marketing team wants to know why their campaign flopped, or the product team needs to understand where users are dropping off, the data analyst digs through all the data and figures it out. Their day revolves around receiving questions from stakeholders, writing SQL queries to pull data, building visualizations in tools like Tableau or Power BI, and writing reports. They need the ability to translate vague business questions into concrete analyses, communicate complex findings to non-technical people, and tell a compelling story with data. This is the most accessible entry point into tech data roles. No PhD required, and you can learn the core toolkit, SQL, Excel, and a visualization tool in a couple months. The skills transfer across industries, which means job opportunities in healthcare, finance, marketing, and retail, not just tech companies. Compensation is still good, but the lowest on this list. The AI impact here is worth taking seriously, though. Tools like Power BI Copilot and ChatGPT can now automate some of the basic analytics stuff like generating visualizations. Job postings declined about 15% year over year. But I am very confident that people with strong business intuition and subject matter expertise will remain in demand. Now, if you prefer more autonomous, self-directed work, let's talk about the path where you get longer focus blocks and you get to define your own problems. At this point, you've got two paths ahead, data scientist and applied scientist. Both give you more autonomy and involve tackling ambiguous, open-ended problems. The question that separates them is about breadth. Do you want to specialize in modeling and analysis? Or do you want to be a jack-of-all-trades who does research, data science, and production engineering? Let's start with the more specialized role, data science. The day-to-day of a data scientist is incredibly varied. Some days, they'll be working in notebooks, doing exploratory analysis and testing hypotheses. Other days, they'll be building a predictive model, say, estimating delivery times for DoorDash. They need strong Python skills, a solid statistics background, machine learning knowledge for building predictive models, and SQL for accessing data. There are important variants within this role. Some data scientists come from an economics background and lean heavily on causal inference and experimentation. Others come from computer science and lean towards machine learning modeling. The BLS projects 34% growth through 2034, making it one of the fastest-growing occupations. But the entry barrier is high. Most roles want advanced degrees or significant experience. The career progression isn't as well-defined as it is for engineers, and there's a lot of time doing annoying tasks like data cleaning. So now if you're thinking, I want to do all of that and own the research and production engineering side, let me introduce you to my role, Applied Scientist. Applied Scientists are the ultimate generalists. They combine the analytical rigor of data scientists, the engineering capabilities of machine learning engineers, and the innovative mindset of researchers, but with a focus on practical, production-level impact. The day-to-day is incredibly varied, which is one of the role's biggest draws. Some days, you're more in data science mode, working in notebooks, analyzing data, and prototyping models. Other days, you're building production systems. And you dedicate time every day to learning, reading papers, taking courses, and researching new approaches, since this role is expected to be able to develop novel approaches when needed. The exact balance of research, engineering, and data science varies significantly between teams. Some Applied Scientists lean closer to research scientists, others toward production engineering, and others toward data science. But the common thread is breadth. You need to be a jack-of-all-trades. This role requires the widest skill range of any on the list. Strong production-level coding in Python and frameworks like PyTorch and TensorFlow, deep machine learning and statistics knowledge, comfort with ambiguity, ability to communicate with non-technical stakeholders, and the capability to take your own research all the way to production deployment. Because of this, the job market is more specialized than for data scientists or machine learning engineers. Fewer positions exist, but also fewer qualified candidates. You typically need a master's or PhD, significant professional experience, and potentially published papers. This role is a great fit if you enjoy navigating ambiguity, are comfortable with both coding and math, want to see your work impact products used by millions, and are self-motivated enough to drive projects with minimal direction. It is not the right fit if you prefer well-defined problems, want to focus exclusively on either coding or theory, or if you want a predictable routine. For many people, like myself, Applied Science is a role you grow into over time. Many, if not most, Applied Scientists started as data scientists or machine learning engineers. Your career is a journey, not a fixed destination. And it's important to remember that the boundaries between these roles are blurry, especially at smaller companies where one person might wear three of these hats in a single day. That's not a bad thing. It actually creates natural pathways to transition between roles. Ultimately, there's no wrong answer here. Every role on this list pays well, has strong long-term demand, and lets you work on interesting problems. The question isn't which role is best, it's which role is best for you. And I also have roadmaps on pretty much all of these roles, so check those out up next as well.