From autonomous GenAI agents that think and act like your smartest analyst, to cleaner, more capable functions that simplify complex logic, every feature is designed to move you from manual to magical.

We’ve also added new ways to organize and protect your work with effortless sorting and reordering in the format tool, while base64 encoding keeps sensitive data secure as it moves through your workflows.

Here’s what’s new this month:

Agentic matching built for messy data

Every dataset has near matches that don’t quite align names, IDs, or values that look the same but aren’t identical. Those small inconsistencies slow analysis and break joins.
​

Fuse Agent handles that ambiguity for you, using agentic matching to detect similar values and score how closely they relate. It turns messy, inconsistent inputs into clean, usable connections.
​
Fuse Agent solves for the reality that “perfect data” doesn’t exist, a challenge found almost everywhere outside of test environments.

Agentic structuring for unstructured data

Unstructured PDFs are everywhere invoices, contracts, reports and turning them into usable data is still tedious and inconsistent. Valuable information stays locked in pages instead of flowing into your analysis.
​

Vision Agent solves the problem of inaccessible data hidden inside PDFs. It transforms those unstructured documents into structured, analysis-ready outputs that flow naturally into your workflows.
​

It removes one of the biggest blockers in working with real-world information, transforming static documents into dynamic, usable datasets.

Agentic data transformation and enrichment

Data enrichment often slows when it meets the limits of manual setup predefined logic and static references that can’t adapt as data changes.
​

Infer Agent makes that process agentic. It applies GenAI to perform enrichment and transformation intelligently across datasets, standardizing values, completing metadata, and aligning fields with minimal effort.
​

It marks the shift from static configuration to adaptive, AI-driven transformation built to evolve with your data.

Building knowledge from every analysis
​

Documentation is essential to analytics, but it’s rarely kept current. Stylus automates it, generating clear records of each analysis and what it accomplishes.
​

Each document becomes part of a growing knowledge base that captures how your systems and data work together. Over time, Stylus builds the context that future intelligence can draw from, turning every documented process into shared memory.
​

It keeps your work understandable today and foundational for what comes next.

Turn arrays into clean, readable text

When working with datasets, arrays are everywhere, but when it’s time to share or display that data, it often shows up messy: brackets, commas, and unreadable formatting that takes time to clean.

That’s where the enhanced To_Text function steps in.
It now supports array inputs directly, turning lists into clean, human-readable text in a single step. No more concatenation formulas or manual parsing, just smooth, consistent text conversion every time.

Whether you’re preparing data for reports, merging fields for summaries, or cleaning up exports for easier review, this update keeps your results readable and ready to share.

Organize columns in one click

We’ve all been there, a dataset that’s almost perfect, except the columns are out of order or the sorting is just off. Rearranging them one by one takes time and breaks your flow.

Now, the Format tool gets a productivity boost with quick actions for sorting and reordering. With just a click, you can tidy up your tables by bringing columns into logical order, grouping related fields, and keeping every dataset structured and consistent.

It’s a small touch that makes a big difference: fewer clicks, less manual work, and more time to focus on the analysis itself.
​

Sometimes, sharing or transmitting data isn’t as simple as it seems. API payloads, authentication tokens, and structured fields often need to be encoded for security and compatibility, but handling that manually can slow you down or introduce errors.

The BASE64_ENCODE function converts any UTF-8 string into a clean, Base64-encoded value. It’s designed to make secure encoding effortless right inside your workflows. Whether you’re passing data between tools, preparing API requests, or embedding sensitive fields, this update ensures every string stays safe, structured, and transmission-ready.

No extra steps. No external tools. Just built-in confidence every time your data moves from one system to another.

From optimized workflows to precise analytics, these updates are designed to accelerate your journey from data to impact. Stay tuned for even more powerful tools and clearer pathways to achieving your best work.

Questions? Ideas? We’d love to hear from you.