How enterprise teams want to use AI: Upskilling for insights
With just a spreadsheet and no coding experience, I set out to uncover the future of AI in our company — here’s what I found.
The initial challenge: Turning data into opportunity
All that stood between me and the insights I wanted was a spreadsheet with qualitative user feedback from our company-wide Gen AI challenge. We asked thousands of non-technical employees to share their AI aspirations, and I was determined to understand their hopes and dreams. But as a product strategist who recently transitioned from 12 years of UX design, I had no experience with the coding needed to analyze that spreadsheet.
Instead of seeing this as a barrier, I viewed it as a chance to expand my skills. My experience in UX design equipped me to think critically about user needs, and I was eager to apply that perspective to data analytics with AI’s assistance. Forbes recently shared strategies on how to use generative AI to enhance your career and professional growth, emphasizing the need to embrace this technology. Since I’ve already integrated AI into my work, applying it here felt like a natural next step.
Turning to AI for guidance: The first steps
Our data analysts team was stretched thin, so my very first step was enrolling in an online bootcamp that introduced me to Python, Jupyter, and key data analytic concepts. My goal wasn’t to master data analysis overnight, but to gain enough understanding to get started. After 2 days of setting up my environment and hands-on exercises I took a break from the bootcamp and turned to AI to help me dive into that spreadsheet with employee feedback from our challenge that I anonymized. ChatGPT wasn’t just a fallback, it was a strategic tool that empowered me to navigate a real world scenario, confidently using Python for the first time. Every step, from importing libraries, cleaning data, and troubleshooting was a learning experience that demystified what had been happening behind the curtain this whole time.
“I don’t know anything about using Jupyter or Python. I want to conduct data analysis on a CSV, and so I want you to walk me through step-by-step in very intricate details what I need to do that would have me bringing in the file, importing the right libraries, doing data cleanup, and then ultimately I want to have topic analysis on the top 50 trends in the data from the opportunity column” — My first prompt
Extracting insights: Turning data into strategy
The raw data I was pulling from my notebook with Python didn’t tell a story — yet. By pasting those outputs into ChatGPT, it helped me illuminate the new ideas, pain points, and opportunities our employees were sharing. It synthesized the top ideas for leveraging Gen AI at work, identified clear justifications for why these ideas mattered, and correlated prompts that, if implemented, would help our employees achieve their goals.
“now let’s go into analysis here with you based on the top trends in the opportunity column. turn this text into something succinct and meaningful that can inform our product development, keep it relevant to enterprise knowledge workers, omit anything nonsensical, only refer to the content below, do not include any outside content.”[pasted trends]— One of my prompts
It was exhilarating to see the insights emerge as I leveraged the strengths of both tools: First a machine learning model extracting trending topics as keywords, and then a large language model making it meaningful in various forms. Every time I had a question, I would just write another prompt, eliciting ChatGPT to provide more points of view on the data. In a matter of hours, Python and AI helped me analyze 3 columns of data across 2,387 rows of user feedback on opportunities, ideas, and feature requests—reinforcing everything we’d been hearing anecdotally.
Empowerment through discovery
This journey wasn’t about instant expertise — it was about embracing the unknown, transforming ambiguity into actionable insights, and realizing the power of continuous learning. Extracting something meaningful from the data felt empowering, even if it wasn’t perfect. Stretching beyond my comfort zone and upskilling in new areas was not only possible , it was surprisingly fun and achievable. The knowledge I acquired has given me a deeper appreciation for the tools and processes data analysts use, which will help me collaborate more effectively with them moving forward. It’s an exciting time to work with AI, and I’m eager to continue learning the fundamentals and applying it to the work I do.
Behind the scenes: Screenshots & examples
GPT laid out a clear plan, providing me with 9 steps, complete with code blocks that I could easily copy and paste into my Jupyter notebook.
Step 1. Import necessary librariesStep 2. Upload the CSV file with dataStep 3. Validate data uploadStep 4. Clean the dataStep 5. Clean the dataStep 6. Clean the dataStep 7. Prepare the data for analysisStep 8. Extract topicsStep 9. Interpret and display the top words for each topic
Example Python outputs:
Topic 24: use, information, questions, ask, get, ai, find, users, timeTopic 25: need, information, time, employee, feedback, process, find, manual, maintenanceTopic 26: information, team, time, provide, questions, sales, client, different, document, teams
Now, here’s a sneak peek at some of ChatGPT’s outputs when prompted for prioritized use cases. This was just one of the many interpretations and analyses I prompted GPT to provide, each offering a unique perspective on the data I extracted.
Automated Workflow Generation and Summarization (High Priority)
Use Case: Automating the generation of workflows that summarize key information, create transcripts, minutes, and action items from meetings, and provide relevant summaries from various communication channels (e.g., emails, Slack).Justification: High demand for tools that streamline the documentation and follow-up processes, ensuring that critical details are captured and distributed efficiently.
Enhanced Workflow Support for Project Management (Medium Priority)
Use Case: Supporting project management workflows by automating task approvals, managing customer requests, and integrating data-driven insights to optimize project timelines.Justification: Project management often involves complex workflows that benefit from automation and data integration to keep projects on track.
Document and Contract Analysis with AI (Low Priority)
Use Case: Leveraging AI to assist in the analysis and processing of contracts and business documents, focusing on extracting key information and summarizing terms.Justification: This use case is niche and may not be universally applicable across all knowledge workers, though it is critical for legal and procurement teams.
If I could accomplish this much with Python and AI in just 3 days, imagine what you could achieve. Start your own journey into data analytics today, even if you, like me, are starting at square one. I began learning Python with this bootcamp, and the best way to make the most of AI is by improving your prompt engineering skills. Have fun and experiment, letting your design thinking skills stand out as you shape the data.
How enterprise teams want to use AI: upskilling for insights was originally published in UX Collective on Medium, where people are continuing the conversation by highlighting and responding to this story.
