This has become all too common with lastpass. I migrated out during their last incident.
I wonder how does it keep happening and how do customers still continue.
This resonates a lot. The shift from "was I right?" to "does this actually help people?" changes everything. I've found that the engineers who got promoted fastest weren't always the smartest problem solvers, they were the ones who genuinely cared about the end outcome.
The hardest part is that user focus is sometimes at odds with technical cleanliness. You can ship something inelegant but useful, or elegant but slightly off from what people need. Most orgs mess this up by choosing elegance.
Duckdb is an excellent OLAP db, I have had customers who had s3 data lake of parquet and use databricks or other expensive tool, when they could easily use duckdb.. Given we have cursor/claude code, it is not that hard for lot of use cases, I think the lack of documentation on how duckdb functions -- in terms of how it loads these files etc are some of the reasons companies are not even trying to adopt duckdb. I think blogs like this is a great testament for duckdb's performance!
In 2025, many SaaS companies are racing to define their AI strategy. While there are plenty of grand visions of fully autonomous “agentic” products, there’s also a more immediate, practical solution that’s surprisingly powerful: a chatbot-driven user experience.
At Databrain, we’re building a modern business intelligence platform. Anyone who’s ever used tools like Tableau or Power BI knows how complex these systems can become—so much so that companies offer certifications just to confirm you know where to click to accomplish certain tasks.
And it’s not just BI. Many SaaS products are brimming with features that users often struggle to discover. Customers scour documentation or call support just to learn how to perform a seemingly straightforward action. Why should users be forced to memorize new menus and workflows? Why invest so much cognitive effort into navigating a product?
Enter the chatbot-driven UX: Rather than forcing users to master every nook and cranny of your interface, a chat-based system lets them simply tell the AI what they want. The AI can guide them to the correct page, perform the requested action, and allow the user to verify or tweak the outcome in real time. It’s a natural stepping stone on the road to “agentic” interfaces, where the AI does most of the heavy lifting but still leaves room for human oversight.
I recently built a prototype to illustrate this concept. At Databrain, we plan to launch this feature soon as part of our platform’s evolution. We previously tested an “AI data assistant,” but it didn’t fully catch on—mainly because people want to trust the output, and they get concerned when it isn’t 100% accurate. Now, I’m convinced a “co-pilot” approach, where humans remain in control while AI assists in routine tasks, is the right balance. This chatbot-driven UX is our first step in that direction.
Ultimately, the idea is to minimize the user’s cognitive burden and accelerate the path to actionable insights. We’ll share more about our progress at Databrain in the coming days, but I’m excited to see how this approach changes the way people interact with complex SaaS products.
Looks interesting! Even without voice, it would feel great. Btw, I couldn't do any other actions like scroll or click on other links, it would be nice if there is a way to get back to the demo when needed.
It doesn't return the data as pandas dataframes, I think pandas ai would be a better if you're looking into something more specific to pandas and chat with db.
Just realised there is a key difference based on another comment.
Adding the same again here.
I think it is the approach — especially on error where it fixes on its own. In Dravid for eg if you say “create elixir project” if elixir is not installed, it will try to install on its own and if there is architecture like for eg M1 related it will fix those errors and finally creating the elixir project. There is no hand holding required.
I think it is the approach — especially on error where it fixes on its own. In Dravid for eg if you say “create elixir project” if elixir is not installed, it will try to install on its own and if there is architecture like for eg M1 related it will fix those errors and finally creating the elixir project. There is no hand holding required.
That’s the main difference
Thank you! I have a section for contribution, I want to add more examples on how one can contribute using the tool, they don’t need to do much. They can have an idea or suggestion and the tool can do the work for them.
I want more people to contribute to make it better user experience and solved problem.
@kosolam I haven't extensively used aider.chat to compare. One option that exists in Dravid would be "dev server monitor" part, which listens to the errors as the dev server is starting and then autofixing it. But aider.chat is definitely feature rich with docker, playwright and several things. This (Dravid) tool is a minimal terminal specific coding framework.
@reallymental heheh, maybe I thought I need a realiable AI coding assistant like Rahul Dravid :) And I can say I'm so far relying it suited to my workflow.
Metadata:
For a fresh project that is initiated by this tool, it creates a base metadata file (drd.json in this case) and updates as and when there is new changes.
For a new project, you can run a command `drd --i` which would initialise a metadata based on recursively checking each file content (multiple LLM calls).
Once the metadata file is created, for the subsequent questions it referes metdadata file for the releveant files and loads those file's content into context. For eg, i ask "can you refactor query's main", it would look for "src/query/main.py" and load the content into the context and then respond.
"dev server monitor": I have tried with scenarios when the module doesn't exists, and it installs and rewrites the modules, type error it fixes. It does all the fix each time and does the restart. I may be biased, it seems to work most of the times I tried. Even during the time if its fix is not correct, it gets it in the next attempt. Need to test in more scenarios.
Actually the main command if you try "drd 'create a sample elixir proj'" it will run into error coz elixir doesn't exist and it will try to install and recursively heal until it figures out similar to restart. I thought this is a great usecase, and I have been using this to create some other python projects. Thanks for pointing out "dev server monitor" could be a killer feature!!
"DHH just admitted he stole ideas from #elixir and @elixirphoenix to create ActionCable @railsconf"
Seems like elixirphoenix is on a right path! This tempted me to look into elixir and phoenix.
My Question: How do you think phoenix can save time for a web developer or a team of web developer?
Most of the time, I hear people talking about how fast their language or framework is. I think at the end of the day, all that matters is a tool that saves time (of frustration and redundancy) without any side effects. And you just wrote about one of the pain points -- deployment.