AI-First Data Prep and Blend
Built for Data Teams.
Built for Data Teams.
Cloud & Desktop.
Windows, Mac, Linux or in a browser.
Windows, Mac, Linux or in a browser.
Catalog
Catalog for workflows and data, no matter where they reside.
Everyone has access to the right data and workflows. Tag, document, search, and discover data based on its format, usage, structure, and location.Data Links.
Versioned proxies for files, databases, and APIs, allowing data redirection without workflow interruption. The data schema is inferred based on use within workflows, ensuring uninterrupted data flow, whether it's a CSV on SQL today or a Salesforce query tomorrow.Integrated version history.
Workflows, data files, and data links — versioned and auto-saved every time. Recall what you have done, let others interpret what you have built, and restore any earlier version on demand.Unified data security and classification.
Data Links sharing operates at a granular level, from complete datasets to single query results, and can be governed by classification, such as importance or risk level.Data Links can store encrypted credentials to shield against social engineering and usage of recycled or default passwords, the most common causes of data breaches.
Comprehensive access and entitlements audit.
All user activity is logged, tagged, and searchable, including permitted and denied data access and modifications. In case of an incorrect update or a data leak, you can easily trace its source or revert it to a previous version.Prepare
Ingest structured and unstructured data.
Read and write data from structured sources such as Snowflake, PostgreSQL, SQL Server, Excel, Tableau, and more.Read and write data from unstructured sources such as files and APIs in various formats, including text, JSON, and XML.
Clean and reshape. Ensure data quality.
Address missing values, correct errors, and reshape data by transforming, combining, and filtering.Blend and process data in‑database and in‑memory.
No need to centralize data in a data lake. If data starts in a database, subsequent components run in-database, as far as possible. Otherwise, data is processed in-memory.Analyze
Enso Copilot, an explainable AI.
Chat with your data. Understand precisely how AI computed the results by tracing generated flow graphs and comprehending annotations for each data transformation step.Live, interactive data processing.
Change parameters, see results live. Adding a new component no longer requires re-running the whole workflow — accelerating time to insight.Build and share custom components.
Create and share custom components effortlessly by collapsing graphs, promoting reusable best practices across your organization.Mix Enso, Python, Java, and JavaScript.
Enso compiles all languages to a common representation with a unified memory model. It lets you call methods and pass arbitrary values between them with no wrappers and close-to-zero runtime overhead.Explore your data without accidental data overwrites.
As you develop your workflow, data is written to a temporary location, so you never overwrite your production data by mistake. You can also define multiple lifecycle environments, such as development and production, for every data link.node1 =
shop_locations_parsed
node2 =
shops_revenue_by_day day=(Equal "2023-10-29")
node3 =
join node2 on=[]
node4 =
ai_generated_fn_1 node3
node5 =
Data.read "customers_opinions" format=JSON
node6 =
`py import tensorflow as tf
embedding = "https://tfhub.dev/google/nnlm-en-dim50/2"
model = tf.keras.Sequential()
model.add(hub.KerasLayer(embedding, dtype = tf.string))
# ...
rating, confidence = [], []
df = args[0].copy()
for row in df.rows():
result = model.predict(row["opinion"])
rating.append(result.rating)
confidence.append(result.confidence)
df.insert(1, "rating", rating)
df.insert(2, "confidence", confidence)
return df`
node7 =
Data.write "customers_opinions_with_ratings" format=Infer
No-code or full-code.
One workflow, two equally first-class interfaces. Non-developers build visually; developers read and edit the same workflow as code — keeping every change reviewable and auditable.Deploy
Schedule and monitor workflows.
No expensive "Server" add-on license required. Time-based scheduling of workflows in the Enso Cloud.coming soon
Expose workflows as REST APIs.
Synchronously or asynchronously call a workflow with dynamic parameters, returning any result.Enso
Enso International Inc.
251 Little Falls Drive
Wilmington, DE 19808
251 Little Falls Drive
Wilmington, DE 19808
Product
©2026 Enso International Inc.