Editor’s note: This Zeenea review is based on my hands-on test of the Actian Data Intelligence Platform (formerly Zeenea) on the SaaS deployment, in November 2026. I weighed pricing and packaging, scanner setup and metadata harvesting, the metamodel and lineage, the data marketplace and data contracts, AI features, deployment and security, and how it compares with Collibra, Alation, Atlan, Microsoft Purview and open-source catalogs.
Quick verdict: I recommend Zeenea for governance teams that want a business-friendly catalog with a shopping-style data marketplace, access requests and ODCS data contracts on one knowledge graph, run as SaaS while metadata collection stays on their own servers. The trade-off is no public price list: the only public figure is a $150,000 12-month AWS Marketplace contract for 10 stewards, 3 connectors and 1,000 Explorer users. Dataedo ($18,000 a year) and Collate’s free tier publish an entry price.
Anyone searching for a Zeenea review now lands on actian.com: Actian, a division of HCLSoftware, closed its acquisition of Paris-based Zeenea on September 12, 2024, and zeenea.com redirects to the Actian Data Intelligence Platform. Docs, logins and SSO hosts still use the zeenea name, and the product is unchanged: Studio for stewards, Explorer for business users, a knowledge-graph metamodel and about 85 documented connectors running through a scanner on your own server.
In this review, I’ll take a closer look at Zeenea’s pricing, features, and deployment model, so you can see exactly whether it fits your data governance team or whether a catalog like Collibra or Atlan is the better buy.
Key Takeaways
- Studio and Explorer split the work: stewards curate in Studio, while business users search in Explorer and request access from the item page.
- Data products carry input and output ports, and the Data Contract Builder writes contracts on Open Data Contract Standard v3.1.0.
- There is no public pricing: the one public figure is the $150,000 12-month AWS Marketplace contract with 10 Stewards, 3 Connectors and 1,000 Explorer users.
- The scanner must run continuously on your own Linux or Windows server, and lineage is automatic only for some connectors, otherwise API or manual.
- AI arrived in 2025 and 2026: an MCP server (October 2025), the Data Steward Agent (June 2026) and Ask AI, which is off by default.
Pros and Cons
Pros
- Separate Studio and Explorer apps for stewards and business users
- About 85 documented connectors through one self-hosted scanner
- Marketplace access requests can fire a webhook or API call on approval
- Data Contract Builder on ODCS v3.1.0 with versioning
- MCP server exposes catalog search and lineage to Claude, Cursor and Copilot Studio
Cons
- No public price list; the one AWS Marketplace contract is $150,000 a year
- Scanner needs an always-on server with Java 11 to 25
- Lineage is automatic only for some connectors
- Ask AI is off by default and needs a Support request
Zeenea Pricing: How Much Does It Cost?
Actian publishes no price list for the Actian Data Intelligence Platform: the platform page offers a self-guided tour and a demo request, and the only public Zeenea pricing is an AWS Marketplace SaaS contract. That contract is a fixed price rather than usage-based, and extra capacity is arranged with Actian directly.
- 12-month AWS Marketplace contract ($150,000): for teams buying through AWS. It includes 10 Stewards, 3 Connectors, 1,000 Explorer users and Data Access Request Management. The 24- and 36-month terms carry no public price, and no free trial is shown.
- Custom quote: for everything else, including the dedicated-VPC option for regulated customers and the Federated Catalog, which is a separate paid option.
- Free trial: none published. The self-guided product tour and a demo are the evaluation routes.
| Option | Price | What it includes | Notes |
|---|---|---|---|
| AWS Marketplace, 12 months | $150,000 | 10 Stewards, 3 Connectors, 1,000 Explorer users, Data Access Request Management | Fixed price, no free trial |
| AWS Marketplace, 24 or 36 months | Quote only | Longer contract term | No public price |
| Federated Catalog | Quote only | Domain sub-catalogs in one knowledge graph, per-catalog permissions | Separate paid option |
| Dedicated VPC | Quote only | Dedicated-VPC hosting | Advertised for regulated customers |
The contract counts stewards, connectors and Explorer users separately, so a 4th source connector or an 11th steward moves the quote. Explorer users are the cheap seat: the bundle includes 1,000 of them against 10 Studio stewards.
Is Zeenea Good Value for Money?
Every rival with a published dollar figure comes in far below the $150,000 Zeenea contract, and most of the big names publish nothing:
- Dataedo: $18,000 (Essentials), $24,000 (Data Lineage) and $32,000 (Data Quality) a year, with a 3-editor minimum and unlimited viewers.
- Collate: a free tier with 5 users and 500 assets; Premium and Enterprise are quote only.
- DataHub: Core is free and open source; DataHub Cloud is quote only.
- Microsoft Purview: billed per governed asset per day plus Data Governance Processing Units. Microsoft’s own example puts 200 governed assets at $100 for 30 days.
- Informatica CDGC: billed in IPUs, at 0.83 IPUs per 100,000 daily assets for the first 500,000.
- Alation, Atlan and Collibra: Alation sells a pool of Alation Consumption Units, Atlan prices by edition, cataloged assets and paid seats, and Collibra publishes no price. All three are quote only.
$150,000 buys a catalog for a large business audience: 1,000 Explorer users served by 10 stewards. A team with a handful of stewards and few readers pays far less with Dataedo, or starts free on Collate.
Reviewer’s Notes: Price the contract around source connectors first, because the marketplace bundle includes only 3. Decide whether you need Federated Catalog and the dedicated VPC before either enters the quote, since both sit outside that bundle.
My Experience With Zeenea
Zeenea splits the work between a scanner on my own server and two SaaS apps, so my test started on a Linux host rather than in the browser.
Installing the Scanner and Connecting a Source
I downloaded the scanner from Administration, then Scanners (an API key with scanner permission works too), and installed it on an x86-64 Linux server; Windows is the other supported host. The minimum spec:
- CPU: 4 cores
- Memory: 4 GB RAM minimum, 8 GB or more recommended
- Disk: 20 GB
- Runtime: Java 11 up to Java 25 LTS
The scanner runs as a managed service and must stay running. It opens outbound HTTPS on port 443 with an API key, so no inbound port is needed, and source credentials stay on my server instead of the admin screens. Since scanner v103, secrets can also come from HashiCorp Vault, OpenBao or Google Cloud Secret Manager.
Next I added the source’s connector plugin, wrote a connection file with its name, code and connector ID into the scanner’s connections folder, and restarted the scanner. The connection then registered on the platform under Administration, then Connections, where scheduled jobs run nightly by default; one connection binds to one scanner. DataHub Cloud’s in-VPC ingestion agent and Alation’s hybrid agent follow the same pattern. Connectors for SAP and Oracle applications need Safyr, a separate product from Silwood.
Designing the Metamodel and Curating in Studio
As Super Admin, I opened Studio, then Catalog Design, and edited the Physical & Logical Metamodel templates for datasets, visualizations, data processes and custom item types by dragging properties from the property library into sections. A property can sit only once per template, and removing one leaves its indexed values in place until a bulk delete. I then configured the Glossary Metamodel, linked glossary item types to physical ones, and defined responsibilities such as Data Steward and Data Owner.
As a steward, I imported datasets from the connection through an import path that picks which discovered items enter the catalog, assigned contacts and documented items. Bulk edits and .xlsx imports up to 50 MB carry the volume work; each spreadsheet needs a key column and one item type. I also switched on the Data Steward Agent in Administration, then Agents, and it laid out its description and owner changes as a plan, waiting for my approval before applying any.
Searching and Requesting Access in Explorer
Explorer is the business-user side: keyword search, topics, and item pages with description, lineage and contacts. Ask AI adds natural-language search with an AI summary, but only after Actian Support enables it and only in topics tagged AI Ready.
On an item linked to an access request policy, I wrote a request from the item page with a free-text reason. The owner gets an email and approves or rejects in Studio, the requester is emailed the outcome, and an approval can fire an email, webhook or API call to a ticketing tool that sets the real permission. Collibra also pitches a self-service data marketplace, while Zeenea keeps the request on the same item page the reader already opened.
Reviewer’s Notes: Design the metamodel before the first import, because templates and responsibilities decide what stewards fill in. Give the scanner host 8 GB or more of RAM before connecting many sources, since every connection binds to one scanner.
Metamodel, Connectors and Lineage
Everything in Zeenea sits in a knowledge graph whose metamodel I shaped in Catalog Design: physical and logical item types (datasets, visualizations, data processes, custom item types) plus a separate glossary metamodel with its own links to physical items. The cost of that freedom is setup work, because stewards see consistent fields only once the templates are done.
Connectors and the Scanner
The docs list about 85 connectors, 6 of them marked deprecated, delivered as downloadable plugins with a minimum scanner version each; the marketing page claims 100+ connectors and APIs. Examples include Snowflake, Databricks, BigQuery, Redshift, Microsoft Fabric, Power BI, Tableau, dbt and Kafka. Collibra claims 100+ native connectors, so on headline counts Zeenea vs Collibra is a tie, and your own source list decides it.
Profiling and sampling run in the scanner, which shipped these changes in 2026:
- v97 (May 28, 2026): sampling targets only datasets without a sample or with a changed structure.
- v103 (October 21, 2026): separate filters for inventory, sampling and profiling, plus secret-manager support.
- v104 (October 30, 2026): collects data quality and observability information, such as Snowflake Data Metric Functions, for SDK-based connectors.
Actian Data Observability writes quality scores back into the catalog, so quality context sits next to the metadata.
Lineage
Lineage arrives three ways: automatically for some connectors, pushed through the API, or entered by hand in Studio. Field-level lineage is modeled through the Catalog API as operations with input and output fields on a data process, and the diagram highlights field lineage across datasets. Lineage shows on dataset, visualization and data process items, expands level by level and exports to PNG. Select Star, now part of Snowflake, built its pitch on column-level lineage.
My verdict: strong modeling flexibility, connector-dependent lineage. Sources without automatic lineage leave API or manual work for your team.
Data Marketplace, Data Products and Data Contracts
The Enterprise Data Marketplace works like an e-commerce layer on top of the catalog: producers publish data products with input and output ports, consumers explore and acquire them in Explorer, and each output port is tied to a data contract. I built a user-defined product in Studio; synced products come from the Data Product REST API or CI/CD instead.
Data Contract Builder
The builder supports Open Data Contract Standard v3.1.0 and opens from the app switcher in both Studio and Explorer:
- Starting points: SQL DDL, a blank contract or existing ODCS YAML
- Collaboration: guided forms, contributors, suggestions and discussions
- Versioning: version history and version compare
- Activation: validation before a contract goes active
- Lifecycle: Draft, Active, Deprecated and Retired
- Export: YAML
The Data Product Builder also generates an ODCS contract from a product’s output ports, so a contract does not start from a blank page. The marketplace also references the Open Data Product Standard (ODPS) for the products themselves.
Access Requests and Federated Catalog
Access requests (see My Experience above) run on admin-defined policies that cover datasets, visualizations, data products or custom item types, name approvers, and can require a free-text or use-case reason. Approval hands off by email, webhook or API call, and the Access Request REST API reports status back.
Federated Catalog is a separate paid option, activated through Customer Success, that splits one knowledge graph into domain sub-catalogs with per-catalog permissions and extends to the marketplace. Each domain curates its own assets, and connections feed specific sub-catalogs. Item Lifecycle Policies shipped on June 3, 2026, adding draft, review and approved stages with visibility rules. data.world, now part of ServiceNow, also sells a Marketplace product.
For data mesh teams, ports, contracts and access requests on one graph are the strongest reason to pick Zeenea.
AI Features: Ask AI, the Data Steward Agent and MCP
The Actian MCP Server, first announced October 28, 2025, exposes catalog metadata to Claude, Cursor, Windsurf and Microsoft Copilot Studio through tools such as search_items, semantic_search_glossary, search_contacts, get_item, get_dataset_data_model and get_item_lineage. Actian also provides remote authentication and a Claude setup guide for it.
- Data Steward Agent (June 3, 2026): a chat agent in Studio that finds items, writes and improves descriptions, sets owners, links glossary terms, creates items and makes bulk changes. It shows a plan and waits for approval, splitting bulk changes into batches, and stays off until an admin enables it for Data Steward-type groups. Admins can set shared instructions for all users.
- Ask AI: natural-language search in Explorer with ranked results and an AI-written summary built from item descriptions. It is off by default, enabled through Actian Support, and works only in topics tagged AI Ready. It runs on Amazon Bedrock (Nova Pro and Nova Micro) in a dedicated VPC, and Actian says Bedrock neither retains prompts nor trains on the data. It is optimized for English, with limited French and German.
- Chrome extension (Winter 2026 launch, February 24, 2026): glossary definitions and report details inside Power BI and Tableau. The same launch added native Microsoft Fabric integration.
Atlan bills AI separately in Flexi Credits, while the $150,000 Zeenea contract lists no AI line at all.
The AI is only as good as the descriptions your stewards wrote: Ask AI builds its summaries from item descriptions and runs only in AI Ready topics, where most items are fully described.
Deployment, Security and the Actian Takeover
Zeenea runs as SaaS on AWS (Amazon EKS, Aurora, S3, Solr for search, Auth0 for sign-in), with a dedicated-VPC option advertised for regulated customers, while metadata collection runs on my own scanner host, which Actian recommends updating at least every 6 months. Actian does not publish customer-selectable hosting regions, and no customer-managed or on-prem install of the platform is documented, which is where Collibra (on-prem) and Alation (customer-managed) pull ahead.
Security
- Compliance: SOC 2 Type II and ISO 27001, as stated by Actian
- Sign-in: SSO through Auth0 with SAML, and OIDC is also cited
- Provisioning: SCIM user provisioning
- Access control: role- and attribute-based
- APIs: GraphQL for the catalog and catalog design, SCIM, and REST for data products and access requests, all authenticated with an API key created in Administration
What the Acquisition Changed
HCLSoftware announced the deal on August 8, 2024, and Actian closed it on September 12, 2024; Zeenea’s CEO joined Actian’s executive team. The product is now the Actian Data Intelligence Platform (formerly Zeenea), sold beside Actian Data Observability, DataConnect, AI Analyst and Jaspersoft, and Observability writes quality scores into the catalog.
The closing release kept the Zeenea brand intact only “for the immediate term”, and docs, login and SSO hosts still carry the zeenea name, so both names show up in procurement paperwork. Select Star went to Snowflake and data.world to ServiceNow, so two other catalogs on this shortlist changed owners too.
My deployment verdict: buy it as SaaS with local collection, not as a self-hosted catalog.
How Does Zeenea Compare to Competitors?
- Collibra: no public price, cloud or on-prem deployment, 100+ native connectors claimed and a self-service data marketplace. It fits teams that need the on-prem deployment Zeenea does not offer.
- Alation: a quote-only pool of Alation Consumption Units, with cloud, hybrid agent or customer-managed deployment and an Open Connector Framework. In Zeenea vs Alation, Alation fits teams that want customer-managed deployment.
- Informatica Cloud Data Governance and Catalog: IPU consumption, at 0.83 IPUs per 100,000 daily assets for the first 500,000 and 0.32 IPUs per compute-unit hour for scanning. It fits teams already buying Informatica IPUs.
- Atlan: priced by edition, cataloged assets and paid seats, with free guests and AI billed in Flexi Credits. It fits teams with many occasional readers, where Zeenea instead bundles 1,000 Explorer users into a fixed contract.
- Microsoft Purview: pay-as-you-go per governed asset per day plus Data Governance Processing Units; Microsoft’s example is 200 assets for $100 over 30 days. It fits Microsoft-centric estates, though Zeenea added native Fabric integration in February 2026.
- DataHub, OpenMetadata and Collate: DataHub Core is free open source with no SSO out of the box, and DataHub Cloud adds a 99.5% SLA and an in-VPC agent. OpenMetadata is open source, and Collate’s free tier covers 5 users and 500 assets with unlimited connectors. They fit teams that want open source or a free start.
- Dataedo: $18,000, $24,000 and $32,000 a year, with a 3-editor minimum and unlimited viewers. It fits budget-capped teams.
- Select Star and data.world: now part of Snowflake and ServiceNow. They fit buyers already standardized on those platforms.
Zeenea fits best where a small steward team serves a large business audience through a marketplace with access requests and data contracts.
How I Reviewed Zeenea
I worked in the Actian Data Intelligence Platform (formerly Zeenea) on the SaaS deployment. I installed a scanner on a Linux host, added a connection, configured the metamodel in Catalog Design, imported and documented datasets in Studio, searched and requested access in Explorer, built a data contract, and tried the Data Steward Agent and the MCP server. I weighted these areas:
- Pricing and value: the AWS Marketplace bundle against published rival prices
- Setup and scanner: host spec, connection files and secrets
- Metamodel and lineage: templates, glossary and lineage paths
- Marketplace and contracts: data products, ODCS contracts and access requests
- AI features: Ask AI, the Steward Agent and MCP
- Deployment and security: hosting, SSO, SCIM and APIs
- Alternatives: Collibra, Alation, Informatica, Atlan, Microsoft Purview, DataHub, OpenMetadata, Collate, Dataedo and Select Star
Prices current as of November 2026.
Zeenea Review: Should You Use Zeenea as Your Data Catalog?
Yes, if a small steward team must serve hundreds of business users. The marketplace bundle’s 1,000 Explorer users, access requests, data products and ODCS contracts all point at that shape, on SaaS with metadata collection on your own servers and SOC 2 Type II and ISO 27001 in place.
Skip Zeenea in these cases:
- You need on-prem or customer-managed deployment: shortlist Collibra or Alation.
- You want a free or published entry price: Collate’s free tier, or Dataedo from $18,000 a year.
- You need deep automatic lineage on every source, since Zeenea collects it automatically only for some connectors.
- You live in Microsoft Fabric and Purview already covers the estate.
Before signing, settle:
- Connectors: how many the quote includes beyond the bundle’s 3.
- Federation: whether you need the paid Federated Catalog.
- Naming: Actian’s roadmap for the Zeenea name on contracts and logins.
Staff the scanner too: it must run continuously on your own Linux or Windows host, and Actian recommends updating it at least every 6 months.
Start with the self-guided tour at Actian, using one real data product and its approval flow as your test case.
FAQ
Is Zeenea still available?
Yes. Zeenea is sold as the Actian Data Intelligence Platform, after Actian closed its acquisition on September 12, 2024. Zeenea.com redirects to Actian’s platform page, while docs, logins and SSO hosts still use the zeenea name. Actian said it would honor existing Zeenea customer contracts, and the product still runs on Studio, Explorer and the scanner.
How much does Zeenea cost?
Zeenea has no public price list. The AWS Marketplace contract costs $150,000 for 12 months and includes 10 Stewards, 3 Connectors, 1,000 Explorer users and Data Access Request Management. The 24- and 36-month terms are quote only, and Federated Catalog is a separate paid option. See Zeenea Pricing above for the full table.
Is Zeenea SaaS or on-premise?
Zeenea is SaaS on AWS, with a scanner on your own x86-64 Linux or Windows server for metadata collection. Actian advertises a dedicated-VPC option for regulated customers, but no customer-managed or on-prem install of the platform is documented. The scanner must run continuously, and it is a collector, not a self-hosted copy of the catalog.
Does Zeenea have data lineage?
Yes. Lineage is automatic for some connectors, pushed through the API, or entered by hand in Studio. Field-level lineage is modeled through the Catalog API as operations on a data process, and the diagram highlights field lineage across datasets. Select Star built its pitch on column-level lineage if that is your main need.
What are the best Zeenea alternatives?
Collibra and Alation are the closest enterprise alternatives, and both offer deployment beyond SaaS. Informatica CDGC, Atlan and Microsoft Purview bring consumption or seat-based pricing models. DataHub, OpenMetadata and Collate cover open source and a free start, and Dataedo publishes annual prices from $18,000. See How Does Zeenea Compare to Competitors? above for each fit.




