Atlan - Active Metadata Platform for AWS

What do you like best about the product?

- Standardized data contracts using YAML templates
- Compatibility with open source frameworks like Apache Spark, Airflow
- Great technical catalog for data engineers
- Lineage packages and libraries

What do you dislike about the product?

- Not user friendly for line of business users
- UI / UX was perceived to be too technical
- Great technical catalog but doesn't cover business governance requirements like CDEs

What problems is the product solving and how is that benefiting you?

- Data Lineage Tracking
- DataOps Monitoring
- Data Contracts Enforcement

What do you like best about the product?

Our experience with Atlan has been exceptionally positive, from initial exploration and technical evaluation to implementation and ongoing operational support. We included Atlan as part of our data infrastructure due to its intuitive user interface, robust product features, and forward-looking vision. Atlan's compatibility with the rest of our modern data stack (Snowflake, dbt, Sigma Computing) and other enterprise tools such as Slack, Google Workspace, and Chrome was a key factor, too. The Atlan team provided ample opportunity to carry out a Proof of Concept (POC) and Proof of Value (POV), which included live demonstrations of functionalities to the various data stakeholders. Additionally, Atlan's pricing and support model were both reasonable and adaptable.

Setting up the platform was swift, with a few hours of onboarding calls spread over a week. We could quickly run scans on our Snowflake instance and set up integrations with other platforms like dbt. Later, when support for Sigma Computing was introduced, we could integrate the same on our own. Features such as web search of metadata and detailed column-level lineage for root cause and impact analysis were an instant hit with our data power users.

The Atlan customer success team was meticulous in providing us with the necessary support to improve the adoption and engagement on the platform. They consistently strive to understand our specific use cases, deliverables, and business outcomes to provide optimal support. We appreciate how the Atlan team has supported us so far with active metadata management and helping improve the governance of our data and machine learning products.

What do you dislike about the product?

A strong product roadmap backs the vision, but the communication regarding the upcoming features and releases can be better. The concepts of Personas and Purpose overlap, and explaining them to most stakeholders is difficult.

What problems is the product solving and how is that benefiting you?

Helping find information about data is easier in a self service way that reduces dependency on data engineers and helps them focus on writing code. The lineage helps during impact analysis and troubleshooting. Establishing a better data governance is easier with a tool like ATlan.

What do you like best about the product?

The ability to apply definitions and tags for almost any data asset is useful, especially the feature to attach existing documentation as a resource. Great search feature, and ability to propagate tags downstream from objects.

What do you dislike about the product?

The UI is sometimes difficult to navigate, and there are some screens that may be more suitable to view as a table. There's also sometimes a steep learning curve to go from basic read-only use (or single asset management), to enabling mass-tagging/documentation efforts.

What problems is the product solving and how is that benefiting you?

Atlan is supporting our initiative to tag and classify all sensitive data in our source system and downstream applications. Being able to propagate tags and view lineage has been really useful.

We are a consulting company, and we propose Atlan  as one of the tools in the market. We inform clients that if they are considering a data catalog, they can also consider Atlan . We evaluate it for them, determining whether it would make sense for their organization.

The best feature of Atlan is its integration with communication platforms like Microsoft Teams  and Slack , so business users don't have to go into a data catalog to see metadata about data assets. This integration feature is the coolest thing about Atlan. The ML capabilities that suggest data classifications and provide data descriptions are also impressive.

One of the main areas for improvement is its governance capabilities. Atlan supports only basic out-of-the-box workflows, and it becomes challenging to customize features like how data owners should approve access to data assets. Its performance is not optimal when dealing with larger datasets, particularly legacy data assets, as the performance declines when scanning datasets running in terabytes.

I have done a couple of POCs using Atlan while working for a company. We were evaluating a few data catalogs, and we included Atlan as one of the prospects.

During our POC, the recently launched ML classification system was a hit-and-miss. However, the support team acknowledged it and since then, the feedback from various users indicates that the issues have been resolved, and it's now working well.

Atlan integrates well with smaller datasets, making it suitable for agile companies. However, it struggles with performance when dealing with larger datasets, particularly those running in terabytes.

During the POC, we had a dedicated account executive, and we received good support from them. They helped us navigate our challenges and brought in technical resources whenever required. There were instances when responses took longer than expected, but this could be attributed to us not being a full-time paid customer at that time.

Positive

If implementing the cloud instance, the setup is straightforward and simpler than other tools I have experienced. However, placing it on-premises requires support from data engineering or technical associates.

In comparison to established players like Collibra  and Informatica, Atlan is cheaper. However, compared to the next generation of data catalogs like Castor, Atlan is pricier. For mid-sized organizations, Atlan provides a good pricing fit.

We evaluated Atlan alongside other data catalogs when working for a company.

Atlan is an eight out of ten, primarily due to its need for improved governance features. If these features are enhanced, it is a ten on ten tool. 

Atlan has unique ML/AI capabilities that aid engineering teams in documenting without having to start from scratch. 

Additionally, its integration with communication platforms helps users understand context without accessing the data catalog directly.

Hybrid Cloud

Other

One of the primary use cases of Atlan  is as an enterprise data catalog. It takes metadata from multiple types of source systems, such as Tableau  and Google BigQuery . The platform helps surface everything into a unified data dictionary and data catalog.

From a business user's perspective, Atlan  has reduced the need to bother subject matter experts by surfacing all data and context. It has significantly reduced instances of business users contacting technical SMEs with questions about data content, providing time savings from a human hours perspective.

As a senior analytics engineer, Atlan's ability to show end-to-end data lineage is the most important feature for me. It graphically displays the entire data flow, allowing me to understand the flow from source systems to Tableau , including the ability to see SQL scripts behind it and usage metrics. Its capability to automatically pull data descriptions and assign ownership are also noteworthy.

Certain UI changes could make Atlan more user-friendly. While they have an Excel add-in for interaction with Atlan, it's in its early stages and could be improved. Additionally, data observability capabilities could be enhanced to provide more alerts on stale or outdated data via communication tools like Microsoft Teams , Slack , or email.

I have been using Atlan for almost a year now.

Generally, I find Atlan stable. There have been instances where new features temporarily degraded performance. They were quickly optimized by Atlan's engineering team.

There have been no notable scalability issues with Atlan. Its cloud-native backend infrastructure is designed to scale and may require human intervention occasionally, but it handles scalability well.

The customer service and support are pretty good. Any issues are addressed promptly by Atlan's team.

Positive

We looked at Google Cloud  Platform's Dataplex. It wasn't fit for purpose because we needed an active metadata platform that could provide comprehensive data discovery and data dictionaries.

The initial setup of Atlan was conducted as a minimum viable product, starting with basic capabilities and building up. From the user's perspective, it involved training a small community of super users to promote adoption.

The implementation of Atlan required less than five people from different teams, mainly technical personnel.

It's difficult to quantify ROI at the moment due to our immature use of Atlan, however, it has reduced interruptions for technical queries significantly.

The pricing model for Atlan is reasonable compared to other cloud-based platforms. There are different license levels, and we receive a discount on the unit price, making it very competitive.

We evaluated Google Cloud  Platform's Dataplex, among other tools, yet chose Atlan for its comprehensive capabilities in data lineage and metadata management.

I would recommend dedicated resources from the beginning, including a product manager, as it facilitates better planning and adoption of the solution. 

Overall, I rate Atlan a nine out of ten. They continue to innovate and listen to customer feedback.

Public Cloud

Other