Yash Gandhi

OrcaSheets AI Reports - Query data to build dashboards and generate detailed reports

OrcaSheets AI Reports lets you query your data to build dashboards and generate detailed reports. One prompt gives you an executive summary, KPIs, insights and recommendations, ready to share in seconds. Stop building reports manually.

Add a comment

Replies

Best
Mayur Jadhav

Hi again 👋

We're back on Product Hunt, and this time with a feature that I'm most excited to talk about. AI Reports.

Honest confession. Writing reports sucks. I hate it with all my heart.

Running the analysis and putting together a dashboard with cool charts and clever insights is the fun part, but turning all of that into a report you can defend in front of leadership is where it stops being fun.

For me, dashboards show you what happened. But a report shows you what someone close to the work thinks about what happened. That used to take weeks, but now it doesn't.

That's what AI Reports does, and it's one of my favourite things @yash_gandhi3, @navdeep710 and I have shipped. You run your analysis in @OrcaSheets, give it a template for how your team writes things up, and it generates a fresh report on the new data - in minutes.

Here's the thing though. AI gets lost in millions of rows of raw data. But give it a few hundred rows of aggregated data and a template to work against, and it does the job honestly well.

AI Reports is normally a paid feature, but if you're coming in through PH, it's free until next Tuesday 💥 Try it on a report you're already dreading. I'll be here all day to answer any and all questions.

Tell me: what's the report you write every month that you wish you could just generate?

Saul Fleischman

This is a really clever insight about AI's effectiveness with aggregated data versus raw datasets. The constraint you've built in—template + summary data instead of raw tables—makes the output more defensible to leadership, which is the whole point. Curious how your users are approaching template creation initially, or does the system guide them through it.

Mayur Jadhav

@osakasaul We work with teams in converting their current dashboards into templates. We work with them to understand their current pipelines and convert them into repeatable workflows called Recipes. These Recipes bring in structure and predictability as you correctly pointed out. From dashboards to report templates currently is an assisted flow, where we help our customers convert their existing reports into templates using Agents. This helps with smoother transition, rather than inventing something from scratch.

Hiya Chaplot

Monday blues needed a rescue. Here it is :)

Mayur Jadhav

@hiya_chaplot1 right? those pesky Monday reports got nothing on us 💪

Natalia Iankovych

What if I have everything in Google Sheets? Will it understand it? Usually AI systems can’t work with Google Sheets.

Yash Gandhi
@natalia_iankovych we are working on a Google sheets connector, for now download the file from Google sheets and load it on OrcaSheets. once you are done you can start asking questions and generating your own AI reports.
Yash Gandhi

Dashboards do one job really well, which is showing you what happened without anyone's opinion attached. That's exactly why they have a ceiling on how useful they can be.

A report does the opposite job. It's written from inside the work, by someone close enough to a project to hold the numbers against what they actually know about how things went. The output isn't what happened, it's what someone responsible for the work thinks about what happened.

The hard part of building AI Reports was figuring out how to keep that perspective in the loop without keeping a human's time as the bottleneck. The answer turned out to be templates. A team writes a template that captures how they think and frame their work, the AI runs against that template on new data, and the separation is what makes the output reliable in a way prompting an AI directly on raw data never quite is.

Genuine ques - where's the actual bottleneck when you write a report? The analysis or the writing?

Ryuta Waku

Congrats on the launch, Mayur. Japan-based founder here.

One Japan-specific thought: for analytics/reporting tools here, the blocker may not be “can AI generate reports?” but whether it works with messy spreadsheet reality: Japanese column names, mixed JP/EN labels, CSV exports from local tools, Excel-heavy workflows, and privacy-sensitive sales/client data.

The strongest local angle I’d test first is not just “generate reports from data,” but “turn messy Japanese spreadsheets and local datasets into shareable executive reports without sending data to the cloud or waiting for an analyst.”

That could matter a lot for SMB operators, agencies, and solo founders here who live in Excel/Sheets but do not want a full BI stack.

Mayur Jadhav

@wakuta Ryuta, thank you for your response. You've described the wedge precisely. Most analytics tools assume the data is already clean, in English, and cloud-safe. The reality for a lot of teams is mixed-language columns, CSV exports nobody fully trusts, and sales data legal would rather keep off a cloud BI vendor's servers. That's the workflow we're built for. Would love to continue this conversation properly, the SMB and agency context in Japan is one I'd want to understand from someone who's operated there. Mind if I find you on LinkedIn?

Ryuta Waku

@mj_jadhav213 Thanks Mayur — glad that matched the wedge.

Yes, feel free to find me on LinkedIn. I’ll keep the PH thread light here, but happy to continue async there if Japan is something you’re seriously exploring.