Sigma

Go build it. IT approves.

[Sigma @sigmacomputing]

3. Pixel-perfect reports.

Define a report once, generate a version for every region, customer, or tenant, and review changes in a pull request, all with the same governance, data model lineage, and live queries as the UI.

[Sigma @sigmacomputing]

Today, workbooks and reports as code are generally available in Sigma.

Your team can now build, edit, and ship Sigma workbooks and pixel-perfect reports from the AI agent or coding environment they already use. Every one runs live on your data platform under the permissions you’ve already set.

We’ve already seen 1,000+ customers put them to work in a few different ways: https://t.co/P83Jpd5sqe

[Sigma @sigmacomputing]

2. App building.

A coding agent can now build a complete AI app in Sigma from a prompt with workbooks as code. The demo app shown here has scenario modeling, plan review and approval, and a Sigma agent that creates plans and takes action, with write-back to the warehouse.

[Sigma @sigmacomputing]

3. Pixel-perfect reports.

Define a report once, generate a version for every region, customer, or tenant, and review changes in a pull request, all with the same governance, data model lineage, and live queries as the UI.

[Sigma @sigmacomputing]

We’ve been cooking up something big.

Come get the first look:
https://t.co/zp6pXbPcaM

[Sigma @sigmacomputing]

📊 Turn your Sigma Public workbook into a trip to Napa Valley 🍷

We're sending one builder and a guest to Napa for 4 nights at a 4-star hotel, with round-trip flights, a winery tour, and a curated @Wine_com box.

How it works: Post a workbook you authored on LinkedIn, Reddit, or X with a link to it and #SigmaGiveaway, then log the share in Sigma Public.

Enter here through November 15: https://t.co/SE9d3r0zjY

[Sigma @sigmacomputing]

Get connected, get configured, and start building.

Join us on November 11 for a hands-on livestream that will depict how the Sigma MCP server and Sigma CLI work, and how to use both accelerate AI building in Sigma.

Save your spot: https://t.co/m8liZuovCq

[Sigma @sigmacomputing]

New in Sigma: The @motherduck connection is now in public beta and available to all Sigma customers 🦆

MotherDuck gives data teams a fast, simple place to run analytics. In Sigma, you can query that data live to build apps, data models, and reports with the governance and depth the rest of the business needs.

Learn more about what your team can do with MotherDuck data in Sigma: https://t.co/DDGUqFkKuT

[Sigma @sigmacomputing]

One Claude prompt → a working Sigma data model

[Sigma @sigmacomputing]

Sigma Public builders: Our final challenge qualifier is live, and this month's job is managing advertising budgets.

Campaign Control gives you the challenge of managing ad budgets across multiple channels and spreadsheets. The finance and marketing teams are struggling to keep up with new requests and track ROI. Your job is to build a Sigma app that helps the team stay on top of new requests while making smart decisions about what to invest in next.

Submit by 11:59 p.m. PT on November 23 and you could claim the last spot to our Builder Challenge at Workflow 2027.

Happy building: https://t.co/eYV253ncrb

[Sigma @sigmacomputing]

Sigma Tables are now in public beta and available to most Sigma customers.

A Sigma Table is an editable table you create from Sigma that lives in your data warehouse. Any Sigma app or outside system can read from and write to it with the right permissions, so teams update a number once and every app that depends on it sees the change.

Instead of finance, ops, and sales reconciling separate spreadsheets by hand, Sigma Tables give every team one live, governed source of truth.

https://t.co/cuAH8j8rAn

[Sigma @sigmacomputing]

We're excited to announce the winners of the latest Builder Challenge Qualifier on Sigma Public 🏆

Join us in congratulating our 1st-place winner, Manoj, who will advance to the next round at Workflow 2027!

This month's challenge revolved around running a candy factory floor. Our judges loved Manoj's build for its realistic flow from receiving data to taking action. His use of manufacturing theme and maintenance alerts also caught the panel's eye.

See all of the winners and submissions here ➜ https://t.co/PHOFiHNJxK

Want to grab the last spot in the Builder Challenge? Our final round kicks off on Monday, Nov 2.

[Sigma @sigmacomputing]

Turn your Claude prompts into business applications connected to your enterprise data.

Watch and learn on November 4:
https://t.co/lNPcY0qfre

[Sigma @sigmacomputing]

Sigma just earned PCI DSS Level 1 compliance, the most rigorous validation a service provider can earn under the Payment Card Industry Data Security Standard. 💳

All the more reason why banks, insurers, and major retailers run their critical data workflows on Sigma.

https://t.co/I6Lt3f91fM

[Marwan Mattar @marwanmattar]

There's a lot of discussion about the capabilities of AI models. Considerably less attention is paid to the efficacy of the systems we build around them.

Recent work from Hugging Face illustrates just how much this matters: the same model achieved 33% versus 62% task accuracy depending on the agent harness. Optimizing model selection in isolation is insufficient. A cheaper model doesn't necessarily translate into a cheaper or faster system.

At Sigma, we invest heavily in harness engineering and rigorous evaluation, continuously optimizing for accuracy, latency, and cost.

Our latest work on Sigma Assistant is a great example. By rearchitecting our build agent, we improved accuracy from 71% to 86%, completed builds in a third of the time, and reduced inference costs by 3x on complete application builds.

Two things are particularly important to me:

First, these efficiency gains translate directly into savings for our customers, who pay for their own AI inference.

Second, improvements of this magnitude require a mature evaluation discipline that allows us to experiment with different architectures, measure their impact, and confidently ship changes. For this comparison alone, we evaluated both architectures across 130 tasks in 30 evaluation suites, with three trials per task, spanning incremental edits through complete dashboard and application builds.

Better models will continue to raise the ceiling of what's possible. Better harnesses—and the evals that guide their development—determine how much of that potential we deliver to customers.

Great work by the Sigma Assistant team and a fantastic write-up by Halden Lin and Christopher Mouri!

Sigma research: https://t.co/4l8uWeOs3a

[Mike Gerardi @MikeGerardi0527]

Git 🤝 Sigma https://t.co/MAGstMWnWN

[Sigma @sigmacomputing]

📊 Turn your Sigma Public workbook into a trip to Napa Valley 🍷

We're sending one builder and a guest to Napa for 4 nights at a 4-star hotel, with round-trip flights, a winery tour, and a curated https://t.co/yCuL6pbmSg box.

How it works: Post a workbook you authored on LinkedIn, Reddit, or X with a link to it and hashtag #SigmaGiveaway, then log the share in Sigma Public. Each workbook you share on each social platform earns you an entry.

Enter here through November 15: https://t.co/qImKVG74Q2

[Sigma @sigmacomputing]

Sigma Assistant now builds apps 3x faster at a third of the cost ⚡

Accuracy on those builds also increased from 71% to 86% 🎯

Because Sigma Assistant runs on the AI provider each organization configures, the savings show up directly in what customers spend on tokens.

https://t.co/aJuxg5FlKW

[Sigma @sigmacomputing]

We've got a few days left in our latest Sigma Public Builders Challenge!

This month's competition revolves around running a candy factory floor: Batch Control gives you three connected tables covering equipment ranges, telemetry readings, and production and quality records for every machine and shift. Use one or all three to build an AI app with an agent that gives a plant team a way to catch trouble in time.

Prizes include a $100 gift card, Apple AirPods, and a trip to San Francisco to compete onstage at Workflow 2027, our annual conference for Sigma builders.

See the full rules and be sure to submit by Oct 26: https://t.co/b0cDSN2ZHf