Unlock AI Outcomes with NeuralMesh AIDP

WEKA’s NeuralMesh AIDP unlocks AI outcomes from enterprise data — without the complexity of building and maintaining AI pipelines.

With NeuralMesh AIDP, data becomes intelligence.

Data from sources like medical device telemetry, social media interactions, and financial market transactions flow into WEKA’s NeuralMesh AIDP platform.

From there, users can securely query and analyze that data to uncover deeper insights.

Behind the scenes, WEKA manages the AI pipeline — performing data embeddings, vectorization, indexing, and continuous updates.

Users interact with their data through a chat interface or API to gain immediate insights.

Leverage the latest NVIDIA technologies and models all orchestrated by NeuralMesh AIDP.

Let’s walk through an example of a medical researcher querying data and uncovering insights.

We begin by logging into NeuralMesh AIDP as user Alex.

Alex is evaluating emerging Alzheimer’s diagnostics and wants to know more about biomarkers for clinical use. He asks:

“Provide a table of demographics according to the p-tau217 research report by Niklas Mattson-Carlgren.”

Hmm… it doesn’t look like there’s any data in the system on that topic.

Here, NeuralMesh is mounted to mnt/WEKA.

There’s no data in a directory called Alzheimer’s, so let’s copy relevant Alzheimer’s research papers into this directory.

The AIDP system recognizes the change and automatically processes the new files, indexes the data, and adds it to the vector database. The data is now queryable.

Let’s rerun that same session.

Something is still wrong. In the enterprise, security must be preserved from the source document through its vector representation.

If Alex doesn’t have access to the original file, he shouldn’t be able to interact with its vectorized data during a chat session.

Let’s correct the file ownership.

WEKA captures file system metadata at ingest time, including UID, GID, file ownership, and source file system.

This metadata can be used to enforce security and access controls before a query is executed.

Here we can see that the file is owned by the root user. Let’s adjust the permissions to Alex’s UID and GID of 1000.

The system picks up these changes in real time and updates user permissions downstream.

Let’s rerun the query.

There we go – Now we have the results that Alex is looking for.

Next, let’s enrich the dataset with additional research papers.

The AIDP system now runs against a richer corpus of data.

And the results are even more thorough.

Let’s verify that the query results are pulling from the specific files we added to the file system.

To confirm the new files are accessible, Alex asks a broader question:

“How close is plasma p-tau 217 to clinical deployment?”

Let’s take the output from the query results and open a terminal editor to search for the same file reference.

Multiple files were referenced in this query.

It’s as easy to remove data as add it. Let’s remove these reports and requery the database.

A delete event is generated and propagated through the pipeline.

When we refresh the UI, we can confirm that the models can no longer retrieve vectors derived from the deleted files.

NeuralMesh AIDP keeps enterprise data continuously flowing through AI systems, evolving intelligence and unlocking AI outcomes.