Shuo Jia

Meta's recent move to allow AI in technical interviews has sparked a lot of discussion, and it's a trend I've witnessed firsthand in my own recent interviews. I was encouraged to use tools like ChatGPT & Claud, which really highlights a major shift in the industry. This experience perfectly mirrors the findings in my recent research paper, "The Impact of AI Literacy on Employment: Evaluating the Risk of Displacement for Non-AI Knowledgeable Individuals." My research confirms it: the debate isn't about AI replacing human skills. It's about professionals who know how to effectively wield AI gaining a critical edge over those who don't. Just as software transformed the professional landscape in the early 2000s, AI is reshaping our world today. But it's crucial to remember that AI is a powerful tool meant to enhance—not replace—your unique critical thinking, creativity, and decision-making abilities. I believe we're moving from a "know-it-all" to a "know-how-to-leverage-it-all" culture. What are your thoughts? Have you seen this shift in your industry or in your own hiring processes? #AI #ArtificialIntelligence #AILiteracy #FutureOfWork #TechTrends #CareerDevelopment #Hiring #JobMarket #Research

8

1 Comment

𝙒𝙝𝙖𝙩 𝙞𝙛 𝙮𝙤𝙪𝙧 𝙖𝙣𝙖𝙡𝙮𝙩𝙞𝙘𝙨 𝙥𝙡𝙖𝙩𝙛𝙤𝙧𝙢 𝙘𝙤𝙪𝙡𝙙 𝙞𝙣𝙫𝙚𝙣𝙩 𝙞𝙩𝙨 𝙤𝙬𝙣 𝙩𝙤𝙤𝙡𝙨? For years, the bottleneck in analytics wasn’t models—it was missing capabilities. A business user asks for something new: a demand-forecasting variant, a custom feature engineering pipeline, a new anomaly detector. Someone from the data team has to stop, write the logic, package it, deploy it… and only then can the analysis happen. But meta-skills in to a different model. Instead of maintaining a fixed library of analytics tools, an agent can generate new skills when it encounters a gap. It reads the agentskills specification, studies an example, and writes a fully compliant SKILL file that becomes part of its capability set. In other words: the analytics stack can extend itself. From my experience building multi-model forecasting pipelines across hundreds of SKUs, the hardest part is rarely the model—it’s building the surrounding feature engineering and analysis workflows. Agents that can generate those capabilities dynamically collapse that bottleneck. This is where GenAI begins to genuinely merge with analytics: not just answering questions, but 𝙘𝙧𝙚𝙖𝙩𝙞𝙣𝙜 𝙩𝙝𝙚 𝙖𝙣𝙖𝙡𝙮𝙩𝙞𝙘𝙖𝙡 𝙩𝙤𝙤𝙡𝙨 𝙧𝙚𝙦𝙪𝙞𝙧𝙚𝙙 𝙩𝙤 𝙖𝙣𝙨𝙬𝙚𝙧 𝙩𝙝𝙚𝙢. How do you see self-extending agents changing the way analytics teams operate?

38

3 Comments

Following up from my last blog post I wanted to share another approach for using parameters. Parameters are a tool I use most of the time I am looking to add interactivity to my dashboard. I have found that using top/bottom style filters in your data is an excellent way to build trust in your dashboard while helping stakeholders understand what is driving the overall performance of the primary metrics of interest. Check out the blog here and share your thoughts on how you use parameters along with other use cases or questions you have. #parameters #Tableau #datafam #analytics

12

Hi Everyone ! After a year at BAT focusing on Power BI development and driving automation from Excel-based reporting to scalable dashboards, I’m looking to expand my capabilities further into the data stack. My current focus is on building a stronger foundation in: • SQL (query optimization, joins, window functions) • Data modeling beyond BI layer • Understanding ETL/ELT processes and data pipelines My goal is to transition from primarily front-end BI development into a more end-to-end data role, bridging analytics with data engineering. I’m currently exploring learning platforms and hands-on environments that simulate real-world use cases. Would appreciate any recommendations on: 👉 Practical platforms for SQL and data engineering 👉 Resources that align well with industry applications Looking forward to growing in this space and connecting with others on a similar journey.

27

🩺 Data Cleaning & Preparation – MGH Project (Week 2) In my last update, I shared some of the data cleaning steps I took on the Massachusetts General Hospital (MGH) project as a data analyst. However, there’s one key column in the patients.csv file I didn’t cover, "DEATHDATE", and it deserves special attention. 🔍 Issue Identified The DEATHDATE column has many missing or null values. For most patients, this is expected - not everyone has passed away. However, to ensure reliable insights, the column must be clean, consistent, and interpretable. 💡 Why It Matters For the MGH project, the goal is to analyse patient outcomes, length of stay, and readmission rates. A clean DEATHDATE field will enable deeper insights into: → Mortality rates by age, diagnosis, or insurance type → Financial and procedural costs linked to deceased patients 🛠️ Cleaning Steps in Power Query Step 1 — Standardise Date Format Transform Tab → Data Type → Date ✔ Ensures consistent formatting across all rows. Step 2 — Create an “Is_Deceased” Column Add Column → Conditional Column → If [DEATHDATE] = null, then “Alive”, else “Deceased”. This approach enhances data quality, maintains key patient insights, and prevents blindly replacing missing values — providing a clearer view of patient outcomes and refining future mortality analysis. 😊 This is week 2 task under the mentorship of Kaosarat Ibrahim Stay tuned! ✨ #STCMentorship #DataCleaning #HealthcareAnalytics #NHSRoyalLiverpoolHospital #PowerBI #DataQuality #DataDrivenInsights #SkillsToCareer

1

I’ve been thinking a lot about where AI actually fits into day-to-day analytics work. In my experience thus far, it’s not replacing anything, it’s speeding everything up. The biggest unlocks I’m seeing in my own work: • Automating repetitive model setup • Exploring data way faster than I normally would • Cutting down the time from question to insight But here’s what has become clear: AI doesn’t replace structured thinking, it highlights whether you have it. The people getting the most value from it aren’t just using the tools, they: • Ask better questions • Actually understand the business context • Can turn outputs into decisions The advantage still feels like it comes down to how you think. Curious how others are using it in their day-to-day.

14

5 Comments