AI Data Guy

@HarrisFanaroff You can use Quickbooks online and connect it to your AI solution of choice for natural language reporting, ex: "What expenses were between $20 and $40 last February"

What's a good ai-powered bookkeeping service? I don't need a full-time bookkeeper but I do want someone to manage this so I don't have to. Who has a good suggestion for me?

@chrisorzy Same

God damn chandelier BLOCKED MY FACE lol

@toddsaunders Great event. Thank you @toddsaunders

The future of AI is in the suburbs. This is the largest redistribution of software creation in history. And tonight it starts in Westfield NJ.

@toddsaunders Looking forward to it!

The suburbs are the future of AI. We have an INCREDIBLE group of folks coming to this AI meetup in Westfield, NJ This is the first in a series of events we're hosting for people in the area passionate about AI (builders, operators, tinkerers, and the genuinely curious). DM me for details, we are capping it at 50 people.

@toddsaunders @conductor_build @Railway @nikunj Basedash for interacting with structured data

I have all day blocked off tomorrow to try new AI tools I’ve been meaning to dive into. My morning is going to be focused on @conductor_build and @Railway thanks to @nikunj But what else should I try? Give me your top 3 underrated but most used tools.

@bhalligan 8. All database tables with column and table level meta data. Bonus points for a de-normalized data warehouse and intuitive views.

No matter what kind of company you are...start making your internal company data legible to AI. Today. As a founder, you are essentially building two versions of your company: the one humans work in and the digital twin that AI agents navigate to do the heavy lifting for you.

@JamesonCamp Nice work on the dynamic chart, sorry more people didn’t see it 😂

This algo really hates me. I put an hour into writing and making this original content and it has 500 views even though I have 90k followers haha

$645 billion. One year. Five companies. Thats what Amazon, Alphabet, Meta, Microsoft, and Oracle will spend on capex in 2026. Mostly on AI. > The Apollo Program cost ~$290B adjusted. Over 13 years. These companies will spend 2.2x that in 12 months. > The Manhattan Project was ~$36B adjusted. Theyre doing 18x that. > The entire US Interstate Highway System cost ~$634B adjusted over 39 years. They will match it. This year. > During the dot-com telecom bubble, total telecom capex from 1996-2001 was ~$500B over 5 years. Hyperscalers will spend 1.3x that entire 5-year buildout in a single year. We build with Claude everyday, and LLMs are magic, no doubt. And I still cant wrap my head around these numbers. This is either the foundation of the most important technology shift in human history Or the largest misallocation of capital ever Theres no in between

@GamewithDave Jezzball

For anyone who used a computer between 1990 & 2005… what’s the one game you still think about?

@jonathan_wilke I like using Claude Code within VS code

After trying the new Claude Code app for about an hour I just went back to Cursor...

@EXM7777 LLMs are great reading structured data as long as the semantic layer is well annotated. Depends on the shape and structure of the specific data, but a well designed data warehouse performs very well for deep insights.

context management is still LLMs' biggest bottleneck today... are markdown files with structured data and graphs the solution?

@toddsaunders I would love to join

I want to get the most AI pilled folks in NJ for a meetup in Westfield. I have no idea how to source or find these people so please X do your thing!

Thank you for taking the time to think about this. I agree with your point that documents are key to the knowledge layer, and certainly not all stories can be told through tabular data. The logical conclusion is having a hybrid knowledge layer that incorporates both document based data and tabular data. A lot of AI discourse does not take this into account, and focuses primarily on 1) file system documents 2) integrations with 3rd party software which sometimes can be a block box. Data warehouses give control over the table/column naming, grain levels, and linking data across systems. What works best of course will differ by company based on where data lives and the skill level of their data engineers.

Alright I’ve been thinking about this a lot and here are my thoughts: Your counterpoint is thinking about the data warehouse layer. CRM records, financial data, operational metrics across portfolio companies. For that, absolutely the best (btw love the semantic context layer for schema). An agent querying that is precise, token-efficient, and grounded in verified data. I am thinking about the knowledge layer. Documents, institutional memory, the connective tissue between deals and decisions and relationships. For that, linked .md or graph structures are better because the data is inherently relational and narrative, not tabular. Context is king and the more interlinked context the better the intelligence layer will be

"RIP Data Analysts" No... Data professionals benefit the most from advances in AI. Business Intelligence strategy has always been about systems design: pipelines, data enrichment, schema optimization, deep business knowledge. That hasn't changed. AI provides SERIOUS acceleration. I use Claude Code to write better pipelines. I use Basedash to have conversations with my data and generate proactive insights. The 10x exists because of domain expertise PLUS incredible AI tools. Google Stitch doesn't make me an expert designer. Claude in Excel won't make a designer an expert data analyst. Experts are still experts. Now we have superpowers to rule over our domains.

Don’t underestimate the value of traditional relational databases, particularly well structured, denormalized data warehouses in this new eco system. Having a strong data architect that can join customers, products, etc. across multiple systems is more important than ever. Add in an additional context layer at the company, table and column level and it solves a lot (but not all) of issues discussed.

All you’re AI tools do a bad job because the way you have structured the data they’re reading is wrong A must read for anyone who wants to actually implement ai in their business

@sideuspronobiss @JamesonCamp Hadn't heard of it before. Will give it a spin.

@aidataguy @JamesonCamp Have you tried hex? Just tried for the first time couple days ago…it’s kind of scary how good it was

Salesforce is gonna ship everything your favorite AI CRM startup built as a tab in a dropdown menu So will Adobe. So will Intuit. And everyone’s gonna forget those startups existed 30 days later

@thebuggeddev @GoogleAIStudio Very impressive. Is there a "chat with your data" component?

For anyone who believes AI cannot build multi page complex applications, here is my proof. A fully functional 20 page dashboard created with Gemini 3.1 Pro inside @GoogleAIStudio 🔥 It did not take a day or even half a day. It took me less than an hour to complete it in 17 total iterations. The UI design is inspired by @salungprastyo and the rest pages were generated entirely by Gemini following same design system. More features coming soon 🔜

@tearsandguile A use case with real data 🤷‍♂️

@tearsandguile Very nice! A an intelligent chat component would be a great addition for V2.

My boyfriend who is a very good design engineer, just taught me how to use Claude and V0 to create dashboards. We came up with this under 1 hour, and it is mobile responsive too.

@GregKamradt As a former (proud) Excel monkey I am blown away by the leverage provided by these new tools. Claude Code has permanently changed my workflow. The differentiating skills are still the same.. understanding the actual business.

Using Opus 4.6 in Claude Code for data analysis is actually quite insane. As a former excel monkey, I'm doing things that would have taken 2x weeks full time in a prompt. It's doing better than I would have done! The dreaded exec line, "this looks great, but can you show me it again with this small edit?" won't be scary anymore. Now you have more tools to provide critical business insights to your shareholders

AI Analytics is here. Love the detail you go into, but believe there are easier setups using either Claude Code or one of the out of the box AI data dashboard/chat tools. Tremendous opportunity still exists because data modeling and kpi definitions are key to unlocking AI's pwer.

I've never been a fan of Tableau and PowerBI. Static dashboards don't answer dynamic business questions. That's why a new breed of analytics is coming: AI Analytics.

RIP BI Dashboards. Tools like Tableau and PowerBI are about to become extinct. This is what's coming (and how to prepare):