How does the more advanced attachment-reading model (Vision) improve cases Jestor previously struggled to interpret?

In Jestor, AI-powered attachment reading — the Vision feature — moved to a more advanced model, capable of better interpreting complex documents and longer instructions, solving cases where the previous reading failed or returned incomplete data.

The cost of a poor attachment reading

Companies that rely on OCR to log invoices, contracts, or receipts know that reading quality changes everything: a misread field creates manual rework, which cancels out much of the productivity gain from automation.

This kind of failure used to show up exactly in the most challenging documents — off-standard layouts, small print, or more detailed instructions about how to extract each field.

What changed in practice

The attachment-reading model was upgraded to a more advanced version.
The upgrade already applies automatically, with no reconfiguration needed.
Documents with more complex layouts are now read more accurately.
You can give more detailed instructions about what to extract from each attachment.
Usage of this feature remains visible in the usage-metrics area.

Before	After the Vision upgrade
Failed on documents with complex layouts	Better interprets off-standard documents
Long prompts were poorly understood	Understands longer, more detailed instructions
Required manual reconfiguration for hard cases	Automatic update for anyone already using the feature

Why choose Jestor?

Jestor applies this Vision model to the same existing document-reading flow — the user just attaches the file and the system fills in the record automatically, with no manual typing, now with a higher accuracy rate on cases that used to be considered difficult.

Frequently Asked Questions

Do I need to reconfigure attachment reading to use the new model? No, the upgrade already applies automatically for anyone who was already using the feature.

Does the Vision model work for any type of document? It was designed for invoices, contracts, and images in general, with a noticeable improvement on more complex reading cases.

Does this use more AI credits than before? Consumption follows the same usage metric already in place on the platform, visible in the usage-metrics area.

Reference Video

Get to know Jestor

With Jestor, you can automate workflows, connect departments, and build internal systems your way — all without code and with AI support.

Get to know Jestor at jestor.com and see how to take your company's management to a new level of efficiency and integration.

How to build an automation roadmap for your company

A good automation roadmap doesn't start by automating everything at once — it starts with a simple starter template, validated with the team, and expands process by process inside a structured BPMS. That's how Jestor recommends starting any implementation. How Do You Build an Automation Roadmap Step

Common mistakes when automating internal processes and how to avoid them

Among the most frequent mistakes automating processes, you'll find automating a poorly mapped process, granting overly broad access to sensitive data, and editing a live workflow without testing it first. Jestor's BPMS, Security, and App Versioning help you avoid exactly these three mistakes. What Are the

How to measure the ROI of a process automation project

Measuring the ROI of process automation requires comparing the tool's cost against the time it saves and the errors it prevents — not just looking at the monthly fee. Jestor's dashboard helps turn this into concrete numbers, comparing before and after automation. How Do You Calculate the

How to identify priority processes to automate first

Priority processes to automate are usually the ones consuming the most team time, generating the most errors, and most affecting the end customer when they go wrong — in that order. Jestor's dashboard and BPMS help you see these three criteria with data, not guesswork. How Do You Prioritize