No Code Low Code Data Platform
- We need a easy of use tool
- We need a drag and drop functions
- We need “Excel” feature in platform
- Visual represenation of the pipelines
- Easy to set up, one click to launch
- Even business users can build data products
Lets say nothing is easy, but it’s not hard.
Problem with enterprise platform
All that is common questions when I am involved in the Data platform proposal, designing and brainstorming the data platform.
But I would say “It’s lie”.
Everything should by codify, even Excel has Macro and VBA which I do not want to mention here because it is paint…
Approaching the enterprise has a lot of requirement and a SaaS can not adopt those.
… Continue
TODO
- What should be considered when working in the big platform
- SaaS vs PaaS, which best match for requirement of no code, low code
- Data Catalog turns to useless when:
- NO input from stakeholders (they are busy)
- minimal information, causing loss of valuable
- no integrated into existing business processes
- no on-going maintain
In the big engineering platform
You only the chain in whole the system, like you just a member in the team has 100 members, or I can say it is 1000 members.
Every company has their own data stack, but in 2025, it does not matter because the supportive of Gen AI. Using AI we can simply generate the document, code of every component in the codebase (except the SaaS tool, but provider and generator will support that surely).
With Gen AI, you can:
- Gen static web
- Gen the MVC web/app
- Generate the infrastructure
- Generate all the foundation of app / platform…
- Text to SQL
- Image to Text
- Object conversion (images, audio, text,…)
But there is missing part in Gen AI. that is context, domain knowledge, like a SME(s).
At this moment, Model builders are trying to improve the size of entire document of company to train the model in advanced. Tokenerization is not the problem right now.
But like I said, Good data is better than big data”, but today I think **making decision with small data.
Key concept is how to make it as a meaningful, accurate data.
Still coding or writing worth in this situation?
I would say YES.
I made the hard decision is that I am using Vim for writing, that’s making me stun to personalize the blog, article.
You will find and realize which one is writing or generating by Gen AI - which is still good for summarizing, but it’s not right thing to do if you paticipate in the writing anymore.
Using GenAI is the tool, as the assistant and automate other workflow, not writing.
About the SaaS development.
Let’s answer the question Where are you AI/LLM/Gen AI in your application ?.
No right - no wrong to fail at this question.
It’s about use cases, audiences you’re targeting.
Why data catalog ?
In the real world, even the data semantic models were built several times, but they (DE, BIE, DA, …) still create the data catalog and data dictionary by their own. Do you think why?
because they did not it exist right, no portal let they know… and no-one is maintaining it.
I would call it the “documentation redundant”
before we jump into the project, they are working around by creating sheet(s) for sharing… it’s dirty handy with documentation. We need to have some better way to optimize the process. –> Data Catalog, but the key thing is maintain it.
Stay tuned - get hand dirty.