- Build Club
- Posts
- Edition 7: Humans of Build Club
Edition 7: Humans of Build Club
An inside peak inside the arena, spotlighting top ANZ AI Builders, tinkerers and founders
Edition 7: Humans of Build Club
An inside peak inside the arena, spotlighting builders turning their dreams into reality.
Humans of Build Club
[Edition #7] Interview with Quynh Nguyen, with Rebecca Williams, Annie Liao & Clinton Lui.
We are so excited to share the next installment of “Humans of Build Club”. A series spotlighting the authentic human story of ambitious AI Builders.
This week, we’re chatting with Quynh Nguyen, who is currently building an SQL query tool specifically to help data analysts in her team save time writing queries. It will produce queries written in their style, using the company’s dataset and table properties.
![](https://media.beehiiv.com/cdn-cgi/image/fit=scale-down,format=auto,onerror=redirect,quality=80/uploads/asset/file/c77d3d35-7404-4b02-9e09-7f0a87c2545c/SWSX_final_graphics__1200_x_1200_px___11_.png?t=1710451291)
How did it all start? What was the beginning of your entrepreneurial journey like?
Honestly, I don't know. At first, I knew that I was very interested in machine learning. Then I gradually shifted my focus into the AI field. The rise of GenAI provided me with so many learning resources and has been very motivating.
What drew me to machine learning and AI was it’s capacity to help me a lot to automate boring and repetitive tasks more effectively.
I’m thrilled to think about how I could scale my personal solutions up to solve my team’s and other people’s problems as well.
For example, at the moment I am working in data analytics, and I have found ChatGPT really helpful for writing SQL queries. I think it could do much better if it knows about the dataset and the properties of a table. However, I think that ultimately a data analyst should be the one who verifies and makes sure that the query works well.
I see myself supporting this process by making my AI adapt to my style of writing SQL queries. SQL writing specifically for data analysts could be a great value add over existing querying solutions.
What inspires you as an entrepreneur?
I think it's cool! :)))
I have found that AI has significantly optimised my life. Instead of having to do the lower-level tasks that I don’t enjoy so much, I’m able to spend more time on problem solving and creativity.
However, I have also found that many existing solutions are not tailored to my problem well, so I decided to become an AI builder to learn how to personalise AI to further improve my life.
I think my problems can be experienced by other people too, so I would love it if I can make my solutions accessible to and helpful for them too.
How do you acquire your technical expertise for the projects you work on?
I start with ideas.
Then I break it down into tasks.
For each task, I read documentation or watch tutorials.
Then I practice with simple things first.
Finally, I come back to build the idea.
This process depends on my mood and energy levels :))
Can you recommend any books, podcasts, links that have helped you on your builder journey?
Building machine learning pipelines by Huyen Chip. It really helped to raise my awareness around deploying ML/AI systems.
I also love to listen to people in the AI Builder Club, it helps a lot as well.
What are you most excited about right now?
I want to know more about Retrieval Augmented Generation (RAG). One of the biggest problems with AI right now is hallucination, and good use of RAG is one of the most feasible approaches to addressing that.
What are you building right now?
I’m currently working on building a SQL generator that is designed for a specific company’s database. I want to help my team reduce the amount of time we spend writing queries.
My solution will be tailored to the problems our team most care about, and in the style of queries that we write.
Do you have an ask for people reading this post?
I’d love you to reach out to me and respond to this question.
What do you do when you try everything you can find on the internet, but all of the solutions you try fail to improve performance?
My nightmare when building ML/AI systems is when one of the computer vision models I’m using for example stays at the 80% accuracy mark no matter how much I try different solutions.
If you have any ideas for addressing this, please let me know.
You can stay up to date with Quynh Nguyen’s builder journey by following their LinkedIn here
Liked this post?