here's why AI still struggles with data that runs your business and what you can do about it
View in browser
The Insider_Header Images

It’s almost the end of summer — and just like Taylor and Travis, we’re ending the season on a high note: Analytics8 is now a Databricks Select Tier partner. 🎉 More than a recognition of our technical certifications, this changes how we build, what we access, and how fast we move with clients.

 

That’s not the only thing on our minds — today we’re sharing thoughts that have come up on recent convos — from AI’s relational data blind spot to burnout creeping in on high-performing data teams (and how some are pushing through it).

 

Plus:

Our Recommended Read: How to know if you're AI-ready 

Hot Topic: Why is utilizing structured data in AI-powered solutions so hard?  

Emerging Tech Insights: Databricks gets business-friendly; Sigma writes back 

From the Field: Burnout shows up before it blows up 

LOL Moment: OKRs in haiku 

 

Let's get to it!

 

Tracey Doyle

Chief Marketing Officer, Analytics8

Was this email forwarded to you? Subscribe here >

Our Recommended Read 

 

Would your org pass an AI-readiness check?

We’re sharing a real scorecard from a real assessment. If you spot yourself in two or more red flags — missing lineage, fuzzy definitions, rigid pipelines — you’re not ready.

AI Data Readiness_ Results
🔎 Learn how to spot these gaps — and how to fix them before they derail your AI strategy 👉

AI's Blind Spot: Your Rows & Columns 

 

If your AI solution shines when consuming unstructured data like free-form text, images and code, but stalls when trying to use rows and columns in your ERP, CRM, supply chain systems, there is a solution.

 

As our CEO David Fussichen explains, the relational model remains the answer. The next real breakthrough will come from tools that bring relational data models into the world of generative AI.

 

On the engineering side, Joseph Reis lands on the same truth: tools evolve, but structure and discipline (think Clean Code) are what scale beyond the AI demo.

 

The takeaway: If you’re building for long-term ROI on AI efforts, watch the intersection of LLMs and relational models.

Emerging Tech Insights 

A couple tech updates on our radar: 

 

1. Databricks One — Governed Self-Service for Business Users  

Databricks just introduced Databricks One: a simplified, business-user experience for getting insights from data and AI without living in notebooks or SQL. It pulls AI/BI Dashboards, Genie (natural-language Q&A), and custom Databricks Apps into one place.

 

What’s new (and useful):

  • Built for non-technical users: a streamlined UI with curated domains, AI-powered search, and direct access to dashboards, Genie, and apps.
  • Easy, governed access: enable the new consumer access entitlement to onboard business users quickly, while Unity Catalog enforces row/column-level security so people only see what they should.
  • Timing and cost: the full experience enters public beta later this summer and is available to Databricks customers at no additional license fee.

Why it matters: you can finally widen access to insights without standing up a parallel toolset—or compromising governance. 

 

2. Sigma Write-Back: Close the Loop in BI 

Sigma’s native write-back lets teams add new rows, edit values, and manage data directly in the warehouse, not on side spreadsheets. It’s powered by Input Tables, so business users can capture inputs, run scenarios, and drive workflows in the same place they analyze — under full governance.

 

What’s useful now:

  • Close the loop in BI: capture assumptions, notes, and approvals in Input Tables and store them as new tables alongside source data.
  • Plan and “what-if” without exports: enter variables and see downstream impact immediately.
  • Stay governed: row-level access, audit trails, and compliance features preserve integrity and traceability.
  • Scale beyond spreadsheets: write back large datasets without breaking your process.

Net effect: fewer offline files, cleaner handoffs, and a single, governed place to analyze and act.

    From the Field 

    Data Teams are Burning Out 

     

    Everywhere you look: missed SLAs, rework, ghosted deadlines. It’s not a vibes thing — it’s an operations thing.

     

    Data leadership consultant Tony Dahlager spoke with culture and engagement expert Sarah (Kirkendall) Odess about how this manifests and how to fix it:

    • The first sign isn’t bad morale — it’s “I thought you were owning that.” Fuzzy handoffs aren’t just annoying. They’re burnout fuel. Treat it like an operations fix, not a morale talk.
    • Strategy, process, culture: pick your diagnosis. Target misses? That’s strategy. Blocked workflows? That’s process. Silence or finger-pointing? You’ve got a culture problem. Fix the right layer or you’ll chase symptoms for months.
    • Burnout costs. When your best people check out or walk, you’re paying in dropped quality, lost velocity, and a price tag that’s often 2x salary. Use it as your ROI case for investing in team health.

    LOL Moment 

    AI-written OKR haiku:

    Rev Ops

    Quarter OKRs set
    to be determined by fall
    Key result: vibes up

    Have any good data haikus? Send it to our content manager and get featured on our social channels (only if we get woo’d by your words).

     

    Have a great week!

    Tracey

    analytics8-logo-1-768x150

    Transform your business with data.

    LinkedIn
    YouTube
    Facebook
    Instagram
    X

    © 2025 Analytics8. All rights reserved. www.analytics8.com

    Analytics8, 55 E Monroe St, Suite 2950, Chicago, IL 60603, 312-878-6600

    Unsubscribe Manage preferences