Databricks Reviews & Product Details

I like that Databricks is an all-in-one powerhouse where I can do multiple works in one place. It's powerful to manage data from multiple sources and have it in a single UC to manage permissions with row-level security. I also appreciate that I can create experiments, run multiple models, and select the best one from logs, which was difficult on other platforms. Once I learned the setup, it's been easy and comfy to work with. Review collected by and hosted on G2.com.

I find it difficult to use the calculator to determine CPU serving endpoint prices because the documentation doesn't explicitly explain this. It only mentions 1 concurrency equals 1 DBU on the Azure page, which isn't clear. The pricing calculator has a single option for serving endpoints, labeled as medium with four DBU, but lacks separate options for GPU or CPU and their concurrency, making it hard to understand how it works properly. Initially, I also felt it was very tough to learn Databricks and manage deployments of workspaces, although it became easier over time. Review collected by and hosted on G2.com.

What I appreciate most about Databricks is its ability to unify data engineering, data science, and analytics on a single platform. The collaborative environment—especially the notebooks and integrated workflows—makes it much easier for teams with different skill levels to work together without constant context-switching.

Another highlight is the integration with popular tools and cloud services that are widely used in the market today, which makes it easier to move data between them. The performance monitoring and job scheduling features help maintain visibility over pipelines, and the Delta Lake support for reliable data management has also been very useful. Review collected by and hosted on G2.com.

Cost management is one area that could be improved. While Databricks offers autoscaling and flexible cluster options, it’s easy for resource usage to escalate unexpectedly, especially with large datasets and long-running jobs. Keeping costs predictable often requires careful oversight and a solid understanding of the platform’s pricing model.

Additionally, some of the more advanced features—such as fine-grained access controls and more complex job orchestration—can feel less intuitive. The documentation is extensive, but it occasionally leaves gaps that end up requiring trial and error. Review collected by and hosted on G2.com.

We use Databricks daily as our core data platform for building and running pipelines across a medallion architecture, from extracting data out of SAP and Arkieva all the way to reporting-ready datasets. The notebook experience is intuitive, the feature set is massive, and Asset Bundles have made our CI/CD story with Azure DevOps really solid. Integration with cloud services was smooth, and once things are set up they just work. The learning curve can be steep for newer team members, especially around things like Unity Catalog and DABs, and costs can creep up if you're not staying on top of cluster configurations. Support is decent and the docs are strong enough that we rarely need to open a ticket. Overall, it's a powerful platform that does a lot under one roof, and it's hard to imagine our data engineering workflow without it. Review collected by and hosted on G2.com.

The cost can creep up fast if you're not careful with cluster sizing and job configurations, so it takes some effort to keep things optimized. Also, the learning curve for newer team members can be steep, especially around things like Asset Bundles, Unity Catalog, and getting the CI/CD pieces wired up properly. Review collected by and hosted on G2.com.

What I like most about Databricks is how it streamlines the entire data workflow by bringing processing, analysis, and machine learning into one platform. The collaborative notebook environment makes it easy to share code, context, and reasoning with teammates, which helps everyone stay aligned. It also performs strongly on large datasets while abstracting away most of the cluster management, so I can focus on solving the problem rather than dealing with infrastructure. On top of that, centralized access control and clear visibility into data usage support responsible data governance, offering a solid balance between power and ease of use. Review collected by and hosted on G2.com.

Databricks has a few downsides, although many of them feel more like trade-offs than outright negatives. My biggest concern is cost: if clusters aren’t managed carefully, expenses can climb quickly, even though the platform can scale very efficiently when it’s tuned properly. There’s also a real learning curve with Spark and distributed computing concepts, and debugging or performance tuning can be more involved than with simpler tools. Lastly, because it’s a managed service, you give up some low-level control compared with self-hosted systems, but the upside is that it takes a lot of the operational and infrastructure work off your plate. Review collected by and hosted on G2.com.

What I like most about Databricks is that it brings everything into one place, making it easy to work on data, build models, and manage workflows. It helps teams collaborate easily in real time. It also works very fast with large data using Apache Spark, and features like automation and Delta Lake make handling big data much simpler. Review collected by and hosted on G2.com.

One thing I dislike about Databricks is that it can be expensive, especially for large workloads. Sometimes the interface and setup can feel complex for beginners. Also, managing clusters and configurations can take some effort if you’re not very familiar with it. Review collected by and hosted on G2.com.

Unity Catalog has been the single biggest value-add for our enterprise migration. We moved from a Hive Metastore architecture to Unity Catalog and gained centralized governance, lineage tracking, and fine-grained access control across all our data assets without bolting on third-party tools. For a multi-domain organization (finance, manufacturing, supply chain, procurement), having one catalog that enforces consistent naming and permissions across bronze, silver, gold, and platinum layers saved us weeks of manual policy work.

UI/UX: The notebook experience with inline Spark SQL and PySpark, combined with the workspace file browser, makes it straightforward for our team to develop and test transformations iteratively. The SQL editor for ad-hoc queries against Unity Catalog tables is clean and responsive.

Integrations: Native Delta Lake support means we don't manage format conversions. The Azure Key Vault integration via secret scopes (dbutils.secrets.get) keeps credentials out of code. ADF integration for orchestration in our V1 environment was seamless, and Databricks Asset Bundles (DAB) for V2 deployment give us a clean CI/CD path with databricks.yml configs targeting dev/qa/prod without custom scripting.

Performance: Switching to CTEs over temp views in our Gold notebooks reduced cluster memory pressure noticeably. The ability to right-size clusters per environment (1 worker for dev, 3 for production) with Standard_D4ds_v5 nodes keeps costs predictable while maintaining performance for our batch ETL workloads.

Pricing/ROI: The pay-as-you-go compute model paired with single-user security mode clusters means we're not over-provisioning. Consolidating our ETL, governance, and BI serving layer into one platform eliminated licensing for separate catalog, orchestration, and data quality tools.

AI/Intelligence (Genie): Genie Spaces have been an unexpected win. Our business analysts in finance and supply chain can ask natural language questions against curated Gold/Platinum tables without writing SQL. It reduced the number of ad-hoc report requests coming to the data team by giving domain users a self-service path that still respects Unity Catalog permissions.

Support/Onboarding: The documentation is thorough, and the skills-based approach to learning (bundles, Unity Catalog, jobs, SQL) maps well to how our team actually works. Onboarding new engineers to the V2 architecture took about half the time compared to V1 because the platform conventions (medallion architecture, asset bundles, catalog naming) are well-documented and consistent. Review collected by and hosted on G2.com.

UI/UX: The notebook editor still feels behind dedicated IDEs. No native multi-file search, limited refactoring support, and the git integration UI is clunky for teams managing dozens of notebooks across workflow bundles. We ended up doing all real development in VS Code and treating the Databricks workspace as a deployment target, which adds friction. The workspace file browser also doesn't handle folder structures well when you have 50+ notebooks organized by domain there's no filtering, tagging, or favorites.

Integrations: Databricks Asset Bundles (DAB) are a step forward, but the documentation has gaps for complex multi-bundle deployments. We run a shared Global_Utilities bundle that other workflow bundles depend on, and getting cross-bundle references to work reliably across dev/qa/prod targets required significant trial and error. The ADF-to-Databricks integration works, but debugging failed pipeline runs means jumping between the ADF monitoring UI and Databricks job runs with no unified view. A tighter handshake between orchestration and compute monitoring would save hours of troubleshooting.

Performance: Cluster cold-start times remain a pain point for development workflows. Spinning up a single-node Standard_D4ds_v5 cluster takes 4-7 minutes, which breaks flow when you're iterating on notebook logic. Serverless compute helps but isn't available for all workload types yet, and the cost premium is hard to justify for dev/test environments.

Pricing/ROI: The DBU pricing model is opaque for capacity planning. Estimating monthly costs for a project with 30+ scheduled jobs, interactive development clusters, and SQL warehouse queries requires building custom spreadsheets because the built-in cost management tools don't give you a clear forecast by workflow or domain. We've been surprised by cost spikes from jobs that ran longer than expected with no easy way to set per-job budget alerts.

Support/Onboarding: Enterprise support response times are inconsistent. Critical issues with Unity Catalog permissions during our migration took 3-5 business days for initial triage, which stalled our deployment timeline. The community forums are helpful for common patterns, but for Unity Catalog edge cases (cross-catalog lineage, complex permission inheritance), the knowledge base is thin.

AI/Intelligence: Genie is promising but still rough for production use. It struggles with joins across more than 3-4 tables, sometimes generates incorrect SQL against our Gold layer, and there's no easy way to curate or correct its responses to improve accuracy over time. Our business users got excited, tried it, hit wrong answers on moderately complex questions, and lost trust. A feedback loop where domain experts can flag and correct Genie's outputs would make it genuinely production-ready. Review collected by and hosted on G2.com.

As a data engineer, Databricks has become my go-to platform for end-to-end data work. The ease of use is outstanding notebooks, Delta Live Tables, and Genie all have intuitive interfaces that reduce rampup time significantly. Implementation was smooth thanks to excellent documentation and responsive customer support that actually resolves issues fast. I use it daily, and the sheer number of features from Unity Catalog to AI/BI Genie keeps growing. Integration with cloud storage, BI tools, and ML frameworks is seamless, making it a true unified platform. Review collected by and hosted on G2.com.

One challenge is the lack of cost transparency at a granular job level it's difficult to pinpoint exactly which pipeline or notebook is driving up DBU consumption without investing in custom monitoring. Auto scaling clusters, while powerful, can silently balloon costs overnight if not carefully configured with proper limits. Additionally, the SQL warehouse tiers can be confusing to choose from upfront, making budget planning tricky for teams. A built in cost allocation dashboard per job or user would be a huge improvement for day to day cost governance. Review collected by and hosted on G2.com.

Databricks stands out for how well it handles end to end data workflows without needing multiple tools. I can ingest raw data, transform it, and publish curated datasets from the same environment. Features like job scheduling, autoscaling clusters, and Delta tables make pipelines more stable and easier to maintain over time. I also like how version control integration keeps development organized, especially when multiple engineers are working on the same pipelines. Review collected by and hosted on G2.com.

One challenge is keeping compute usage under control, especially when pipelines scale or run more frequently. Without proper monitoring, costs can increase faster than expected. Also, debugging failed jobs can sometimes take time, particularly when dealing with complex dependencies or Spark level issues. The platform is powerful, but it expects a certain level of technical understanding to fully optimize it. Review collected by and hosted on G2.com.

What I like most about Databricks is how it simplifies the entire data workflow. Instead of switching between multiple tools for data processing, analysis, and machine learning, everything is available in one place. The notebook environment makes collaboration really smooth it feels natural to work with teammates, share code, and explain logic without extra effort.

Another thing I appreciate is the performance. Working with large datasets can usually be painful, but Databricks handles it efficiently in the background. You don’t have to worry much about managing clusters or optimizing everything manually it just works most of the time, which lets you focus more on solving the actual problem rather than dealing with infrastructure.

What also stands out is the way it handles data governance and organization. With features like centralized access control and better visibility into data usage, it becomes much easier to manage data responsibly, especially in larger projects. Overall, it gives a good balance between power and ease of use, which is why I enjoy working with it. Review collected by and hosted on G2.com.

One thing I don’t particularly like about Databricks is that it can get expensive pretty quickly, especially if clusters are not managed properly. If you forget to terminate clusters or run heavy workloads without optimization, costs can spike without much visibility at first. For teams that are still learning or experimenting, this can become a concern.

Another downside is that debugging can sometimes feel a bit tricky, particularly when working with distributed jobs. Errors are not always straightforward, and tracing issues across multiple nodes can take extra time compared to working in a simpler local environment. It requires a certain level of experience to quickly understand and fix issues.

Also, while the platform is powerful, it has a bit of a learning curve for beginners. Concepts like cluster configuration, job scheduling, and data governance are not always very intuitive at the start. It takes some hands-on time before you feel fully comfortable navigating and using everything efficiently Review collected by and hosted on G2.com.

Databricks is especially helpful because it brings data engineering, analytics, and machine learning together in a single unified platform, which reduces the need to manage multiple separate tools. Built on Apache Spark, it can process massive datasets quickly and scale smoothly as workloads grow, making it a strong fit for big data use cases. It also supports collaborative notebooks where teams can work together in languages like Python and SQL, which makes it easier for data scientists and engineers to collaborate effectively.

With its lakehouse architecture powered by Delta Lake, Databricks combines the flexibility of data lakes with the reliability of data warehouses, helping ensure better data consistency and performance. In addition, it integrates with tools like MLflow to streamline the machine learning lifecycle end to end, from experimentation through deployment. Overall, Databricks simplifies complex data workflows, improves performance, and helps organizations build scalable data and AI solutions more efficiently. Review collected by and hosted on G2.com.

Databricks does have some limitations, although many of them feel more like trade-offs than outright negatives. A frequently cited drawback is cost: while the platform is flexible and scalable, expenses can rise quickly if clusters aren’t managed carefully. At the same time, that cost often reflects its ability to handle very large workloads efficiently when it’s properly optimized.

Another consideration is the learning curve, especially for beginners who aren’t familiar with Apache Spark or distributed systems. That complexity can be challenging at first, but it also comes with the benefit of powerful capabilities once you get comfortable with it. Some users also find that debugging and performance tuning are less straightforward than with simpler tools; however, Databricks offers detailed monitoring and optimization features that can make these tasks easier over time.

Finally, because it’s a managed platform, there can be a sense of reduced control compared with fully self-managed systems. In return, it removes much of the operational burden that comes with infrastructure management. Overall, while these areas may be seen as the “least helpful” aspects, they’re often balanced by the platform’s scalability, integration, and productivity gains. Review collected by and hosted on G2.com.