What is one thing that frustrates you the most about your data?
by•
Given all the AI tools out there today, what is one challenge with data that still remains unsolved? why do you think it hasn't/can't be solved easily.
267 views
Given all the AI tools out there today, what is one challenge with data that still remains unsolved? why do you think it hasn't/can't be solved easily.
Replies
Bababot
Even with great AI tools, I still struggle to trust my data. There are duplicates, wrong entries, and missing pieces that make me second guess every decision. I wish more tools cared about accuracy instead of just insights.
Querri
@kabir_haynes Correct. We have all heard of GIGO (Garbage In Garbage Out). Dirty data is a source of frustration at so many organizations - both large and small. I read somewhere that companies can lose 15-25% of revenue because of bad/dirty data.
Many data tools, including Querri, now have built-in data cleaning capability, which is helpful, but if you rely on Excel or other manual cleaning methods, the amount of cleaning to be done can seem very daunting. Querri starts each new data connection by auditing the data quality, reporting on it, and then proceeding to clean it
Querri
@kabir_haynes I was just talking with one of our engineers about this. A tricky AI design question is do you answer faster or do you answer more accurately? .. you can improve accuracy by reasoning longer, checking the results in more detail, trying things from more angles, etc. We've tended to opt more and more towards slow and steady to get more accurate answers, but it comes at the cost of people saying "Well ChatGPT gave me an answer faster".
Sounds like you'd be on the side of "take your time just be right"?
minimalist phone: creating folders
That Someone will use my identity to do some things against the law.
Querri
@busmark_w_nika That’s such an important point, most people talk about messy or incomplete data, but the fear of misuse is just as real. Curious, what do you do to protect yourself from this?
@busmark_w_nika @querri +1 this is such a critical issue as a founder where your reputation and its management directly impacts your fundability.
Querri
@busmark_w_nika @querri @john_lunsford_phd absolutely. We became SOC 2 compliant this year and are continually working to keep data safe. It's kind of a must if you want people to trust you with their data.
minimalist phone: creating folders
@querri I usually try to not to use my real emails, use bank cards with €0.05 on them to protect against unsolicited automatic payments etc :D
Querri
@george_esther This is a good point. Which is why the answer to the question "Will AI replace Data Analysts?" is no. An AI-powered data analytics tool can really help speed up analysis, find errors and outliers faster, automate, and more...but company-specific context is best handled by humans in the loop.
Honestly, the biggest frustration is inconsistency. Every dataset seems to have its own logic, format, and exceptions and aligning everything into something usable feels like a never-ending puzzle.
Querri
@sheik_born curious how do you address this today?
Data governance plays a huge part in this. Data is no more a single person or team's responsibility - owning, maintaining, using, securing should be a team effort.
Bababot
My biggest struggle is merging qualitative and quantitative data. There’s so much nuance in feedback and behaviour that numbers alone can't capture. I keep hoping an AI tool will finally bridge that gap.
Querri
@adams_parker That's so very interesting! can you elaborate some more on this usecase.
@adams_parker Have you figured out any way to quantify the qualitative data to help allow you to focus on key parts of data or specific data rather than on larger data sets? From there, then you can dig deeper into the qualitative data for more context.
I think.. What frustrates me most is that data still tells stories without a heartbeat. It shows patterns but hides the people behind them. You can have endless charts and dashboards yet still miss the truth that matters. Maybe that’s the part we have never fully solve because meaning doesn’t live in numbers it lives in the space between them.
Querri
@ravi_mansuriya I think humans should always stay in the loop because context matters ;-).
Querri
@ravi_mansuriya to me, that is a blessing because otherwise humans would become extinct faster than they should :) That's a joke, but seriously, I think we do want to keep that control because that's where humans shine. That's where you data/content will help you stand apart from rest of the crowd.
A lot of the time, from what I’ve seen with engineering data during migrations, the data entered manually is fragmented, poorly mapped, and often doesn’t follow standards. AI can help clean up maybe 60–80% of it, but when you’re dealing with large amounts of critical data, it’s really hard to fully trust AI tools. The real challenge I believe isn’t processing the data — it’s ensuring accuracy, context, and consistency, which are still fundamentally human and organizational problems.
Ideally companies will adopt AI to help with reviewal of newly inputted data to help enhance it. But for the existing data it would be great for an AI platform bto have the ability to flag data issues, run multiple review rounds, use a suite of LLMs to have do group analysis and validate the data from multiple sources even if it takes longer.. Having layered checks could make manual assessment and cleanup far more efficient and reliable. and in my case - give me much more confidence in AI’s capabilities and reliance on it.
Querri
Serand
What frustrates me most is how hard it is to get real-time clarity. My data updates constantly, but dashboards often lag behind. It feels like I’m always reacting instead of leading.
Querri
@rukhsar_amjad what tools do you use to generate your dashboards. We hear this same frustration from many companies. I was surprised to hear that even at large organizations, executives are constantly requesting for screen shots of their latest dashboard from their data teams just a few minutes before the board meeting.
Honestly, the most frustrating part for me is having data scattered across too many tools. It always feels like half the story’s missing.
Querri
@stas_kaufman Every day there are hundreds of new tools created to make our life easier :) And we fall into the trap of using more and more of these tools. The more tools we use, the more data we generate and more scattered it becomes. I am not sure if there is or will ever be a cure to this problem. How many connectors can your data tools build to amass this vast ocean of data.
Tactiq
that it's all over the place. especially with SEO. I'm pulling data from different places.
Querri
@heyitsirenechan scattered data is a challenge for almost every company. Plus, not all platforms make exporting the raw data easy. We've built in more connectors to make data consolidation easier, and added a feature that lets you ask questions straight from your library. That way, you can connect & upload data from various sources and ask questions in one place.