In this video, data‑analytics practitioner Kedeisha Bryan (5 years experience, 500+ placements) explains how recent AI advances are reshaping the analytics profession and the hiring landscape. She cites an MIT study showing AI can complete 50‑75% of analyst‑type tasks at a minimally sufficient level, with projections up to 90% by 2030, and Greenhouse data indicating that the average number of applications per job more than doubled after ChatGPT’s launch. While AI isn’t poised to eliminate analyst roles outright, it has already altered how work is done and how candidates are evaluated. Bryan argues that simply knowing tools like Excel, SQL, Power BI, or Tableau is no longer enough because AI can generate comparable outputs at scale. She introduces a “5‑check AI audit” (total vs source, spot‑check rows, sanity‑check numbers, trace data lineage, assess business meaning) to ensure AI‑generated analyses are accurate and trustworthy. She stresses the need for portfolios that showcase problem‑solving, decision‑making, and error‑catching rather than just polished dashboards. For career changers, she highlights the advantage of leveraging existing domain knowledge (e.g., healthcare, logistics, finance) to add business context to analytics work. Finally, she outlines a five‑step, zero‑to‑hire plan: (1) practice SQL via a whiteboard method, (2) learn basic statistics, (3) use AI daily as a junior teammate and audit its output, (4) build a project that deliberately catches and corrects an AI mistake, and (5) practice explaining findings to non‑technical audiences. The video concludes with calls to action for a detailed roadmap and a follow‑up case study of landing a $100k analyst role in 90 days.
Chapters
An MIT study has just proven AI can do a satisfactory job on between 50% to 75% of appointed tasks without any edits. And there's a prediction that hundreds of thousands of data analyst roles are at risk of being completely replaced by AI by 2030. Now, if you're new here, my name's Kedeisha. I've been in data analytics for the past 5 years, and for the past 4 years, I have been helping career changers move into high-paying data analyst jobs and placing over 500 people in data. So I've seen firsthand the impact AI has had on the job market. And in this video, I'm going to show you exactly what areas of analytics has been affected by AI, where your skillset can still be used, and the exact plan that I would follow right now if I was starting my career over from scratch. So let's get into it. Okay, now before we get into this MIT study, I want to share something from one of my newest students. They were concerned about AI replacing data analysts, and I get it. If you've been spending time and money breaking into this field and seeing all the hype around AI, I understand how scary this may look. MIT researchers tested AI on more than 6,000 tasks and found that it could complete roughly 50 to 75% of what they called minimally sufficient level without any edits. In plain English, all that means is AI's first attempt was good enough to meet basic requirements— not necessarily perfect, but just acceptable. And if AI keeps improving, the researchers project that the number could reach as high as 90% by 2030. Now, those numbers are a big deal, but here's the important distinction. The researchers were testing individual tasks across different occupations. They weren't exactly testing whether AI could replace an entire data analyst job, which is why a lot of these headlines about AI can confuse people. So no, this study doesn't prove that data analyst jobs are gonna disappear by 2030, but that doesn't mean we can ignore what's actually happening either, because whether AI can replace an entire analyst or not, It's already changed how data analytics work gets done today. And I think that's where we need to pay attention to because AI hasn't just changed the work itself, it's also changing how people are getting hired. Think about how people are applying to jobs today. You can use AI to write your resume, tailor it to the job description, write a cover letter, and even automatically submit hundreds of applications a day. That sounds great in theory, except now everyone has access to these same tools. So it's no longer an advantage. In fact, Greenhouse analyzed over 640 million job applications and found the average number of applications per job more than doubled between 2022 when ChatGPT got launched and 2025, going from 116 to 244. That's just giving you an idea of how much more competitive the application process has become, which is why the old approach to breaking into data analytics starts to become a problem. Let's say you spend a few months learning Excel, SQL, Power BI, you complete a certificate, you follow a couple dashboard tutorials, and you put those projects on your resume and start applying to jobs. And you send the same exact resume to 200 different companies hoping someone gives you an interview. And honestly, during the hiring boom of 2020 to 2022, that approach worked for a lot of people. Companies were hiring aggressively, and having those technical skills could actually help you get a foot in the door. But we're not in the same job market anymore because now you're competing against hundreds of other applicants with those same similar certificates similar projects and the exact same technical skills based on their resumes. And here's what even makes it more challenging. The tools that used to help you stand out on their own are now the exact same tool AI has now helped begin to automate. AI can help you write your SQL queries, help you clean your datasets, help you generate formulas, and even help you create dashboards. So just showing an employer that you can produce those same things doesn't necessarily tell them how much you actually understand them. Now, I'm not saying it's a waste of time to learn SQL and Excel, Power BI, etc. What I'm saying is Just knowing them is no longer enough anymore. AI has just made it way too easy to produce things, whether they're good or not. Job seekers can produce more applications and people working on data can produce more analysis and reports much faster. But because something looks complete does not mean the person behind it understands what they're doing. And that's the part that matters, because if you're going to get hired in this job market, you need to be able to demonstrate more than the fact that you can use a few tools. You need to actually show you understand the work. So what exactly does that mean, and what do you need to do differently right now in this modern job market? Let me first show you something a student shared with me. They're working on a project, and they actually used AI to clean their data and write VBA and DAX scripts, the code behind Excel automations and Power BI formulas. Now, everything looked great, and they completed their project, but when they looked back at it, they could not explain how any of it actually got done. And if you cannot explain the work that you're using AI to do, how can you check whether it's wrong or not? That is a skill that I want you to start practicing. Do not just use AI to blindly produce analyses. You need to audit it. Let's call this the 5-check AI audit. These are the checks that I would run before trusting any AI-generated analysis. First, total against the source. Does your numbers agree with the raw data? One time I personally used Gemini inside Google Sheets to analyze a spreadsheet. And the calculations were way off. If I hadn't used my critical thinking and decided to just add the bump myself, I could have used the wrong numbers without realizing. Second, spot check individual rows. AI can write all the code that you want that looks perfect, but it can still be completely wrong. One of my students caught this in a project. Some of the customer IDs had lost their leading zeros, so 00123 turned into 123, and those records stopped matching. There was no error message, and when they asked the AI, it said the code was fine. They only found it because they pulled up a few rows and they checked them by hand. This is a critical check that most people miss, but it will make a difference if you want to work with AI. Third, ask whether the numbers make sense. This is why you still need basic statistics. Is that 200% month-over-month increase real or is there a data problem behind it? You don't need advanced math to ask these questions. You just need those thinking caps that we learned about back in elementary school. Fourth, trace where the numbers actually came from. You gotta understand the basic data engineering behind how information moves from the source to the dashboard or to wherever your report is so you can tell whether the pipeline is sound. And then fifth, ask what it means for the business. A 5% drop in signups might be fine in July, but a serious problem in December. AI can give you the number, but you still have to understand the context behind it. And if you don't know, be humble enough to ask someone else who does. The tool used to be the job, but now checking the tool and checking the AI work behind it is now more of the job. Now, if you wanna prove to an employer that you actually understand the work of an analysis, your portfolio needs to reflect that because having a dashboard in your portfolio is not enough anymore. Employers wanna know how you approach the problem, the decisions that you made, and how you got your final results. That's something a list of tools on your resume just can't show. One of my students who landed a 6-figure data analyst offer told me that having a detailed portfolio made a big difference during their interview. Instead of just talking about their skills or saying that they have a portfolio, They were able to walk through the actual projects on their portfolio live in the interview, and the VP of analytics who was interviewing them loved it. So when we're talking about building your own projects, especially if you're using AI, don't just focus on the finished product or the finished dashboard. Show how you approach the problem, what you checked, any mistakes that you caught, and how you corrected them, because that gives an employer a much better understanding of how you think and solve problems. And in today's age of AI, I'm telling you, critical thinking and problem solving is a much needed skill and honestly hard to find. And if you're switching careers, there's another advantage you might not even be thinking about— your previous work experience. I've coached nurses, radiation therapists, high school math teachers, truck loaders, people with well over 10 years of experience in warehousing and logistics and operations, even 17 years in finance. And while these people might have been new to SQL or Power BI or Tableau, they were not new to understanding how businesses operate in their domain. Think about someone who's been working in healthcare for years. You understand patient care, hospital operations, and the kinds of problems people are working on in that industry every day. So when you're analyzing healthcare data, you can bring that experience into the decisions that you make and the questions that you ask. Because knowing how to write a SQL query is one thing, but understanding what the results actually mean for the business is another. This is why I always encourage career changers to consider industries that they already have experience in, whether that's healthcare care, government, consulting, logistics. Look for opportunities where your existing knowledge can make your analytics skills more useful. That is the number one way to double your income with a new analytics job than anything else. So if you're trying to break into data analytics, don't just ask yourself, which tool should I learn? Ask yourself, what do I already understand that can make my analysis more valuable to an employer? All right, so if I were starting from zero right now, I'd keep the plan very simple. First, I'd learn SQL properly using what I call the whiteboard method, where you basically complete SQL challenges on a Word doc without the luxury of autocomplete or error messages that a code editor would give you. So when you write code in plain English without a code editor, it forces you to think in SQL code, therefore improving your natural problem-solving abilities in SQL. It may be rough at first, but you will thank yourself later when get to the interviews. It's also going to come in handy whenever you're gonna be using AI to help write your SQL code, because if you can't write your own queries by yourself, there's no way you'll be able to quality check whatever the AI helps you write. Second, learn basic stats so that you know when a result just doesn't make sense. Third, use AI every day, but treat it like a junior on your team. It drafts and you check. You are the pilot and AI is your copilot. But think of it like a copilot where you have to go back and always check their work, not where you can just sit back and relax and just go to sleep while it doesn't work for you. Fourth, build one project where you catch AI making a mistake and then document the correction and make that your interview story. And then fifth, explain your finding to someone without a data background. It could be your mom, a friend, an old coworker, or you could even use AI. As a data analyst, you're going to spend majority of your time explaining things to non-technical folks, which is why communication is one of the most slept on skills To become a data analyst. Now notice what this plan does not ask you. It doesn't ask you to collect 5 other certificates. It only asks you to get good enough at the work to recognize when an answer is wrong when you're working with AI. So you're probably relieved to hear that AI isn't exactly taking over your job, but there's still a lot of work required to land a high-paying role in data analytics. If you're wanting a full roadmap of exactly how to land a high-paying data job, click the link in the description. If that's not for you quite yet, then watch this next video where I show you exactly how I landed a $100,000 K data analyst job in under 90 days. See you there.
