Genie code and the inline Assistant were the most helpful tools for me on my project. They helped me debug a 2k-line codebase and clearly explained why I wasn’t getting accurate data. It also provided a query to run in my source system (SQLMI). By running the discrepancy script in parallel on the source and target, I was able to debug the entire code much faster and improve my productivity. Overall, it cut my work time from about 8 hours down to around 1 hour. Review collected by and hosted on G2.com.

In Delta Sharing, there isn’t a catalog-level SELECT permission, and I sometimes think having that would be helpful. Also, when I use the Genie code inside a VM, it can make the website unresponsive at times. These are areas that could be improved. Review collected by and hosted on G2.com.

I appreciate having a dependable source for figures essential to documentation, like installation records, site performance data, and warranty data that I can query directly rather than relying on outdated spreadsheets. I really like the Genie feature because it allows me to ask questions in plain language and get usable answers without waiting for someone else to process a request. It works quickly and can handle large record sets efficiently, which still impresses me compared to previous methods. The dashboards offer a standing view of frequently cited data, which helps me quickly assess any changes rather than manually re-checking each number. The initial setup of Databricks was quite easy as we didn't encounter any major issues. Review collected by and hosted on G2.com.

Data does not always travel between platforms as cleanly as I would expect. Where a table comes across from another system, it is not consistently available to me on this side, so I am left reconciling definitions that ought to match already. It is the one area where I feel the groundwork was laid later than it should have been. Review collected by and hosted on G2.com.

Databricks simplifies the process of running data transformation operations on massive datasets. Although it can be a bit of a paradigm shift from classic asynchronous processing architectures, it is extremely easy to get started with. Simply put, the thing I like best about it is it's ability to do work at scale. Review collected by and hosted on G2.com.

The inability to run a copy of Databricks locally to test changes before deploying them to production is a significant hindrance. Creating per-developer staging environments might be a close solution l, but might be a lot of work to manage. Review collected by and hosted on G2.com.

I love that the Databricks notebook environment makes it easy for our data scientists and engineers to work in the same space without constantly handling code back and forth between separate tools, which used to slow down every model iteration. The integration with our streaming infrastructure has been genuinely solid, allowing us to pull transactions in continuously without a lot of custom plumbing on our end. I also appreciate that pushing score results back out to our decision engine is just as smooth. The performance and load have held up well during peak shopping periods when transaction volume spikes hard, and the pipeline manages to keep pace without falling behind. Review collected by and hosted on G2.com.

What I don't really like is the pricing for running continuous streaming workloads, which adds up fast compared to the batch jobs we used to run since clusters need to stay active rather than spin up on a schedule. Justifying the ongoing cost took some real modeling on our end before finance was fully onboard. The AI assistant built into their platform helps with general coding tasks but doesn't offer anything specific to fraud modeling itself, so any real intelligence in our detection logic still comes entirely from our own data science teamwork, not the platform. Review collected by and hosted on G2.com.

I especially like the ability to generate visualizations within the same platform, which saves me a lot of time by not having to constantly export the data to make the information understandable. The real-time collaboration capabilities and processing speed are undoubtedly excellent, and the fact that it can be installed on a laptop and allows me to view the team's progress is superb. The initial setup was surprisingly efficient. Review collected by and hosted on G2.com.

I believe that, although I am comfortable using IT tools, the user interface for quick report generation could be improved, making it more intuitive for people who are not used to writing code. In other words, the module should be a little more flexible, allowing dragging and dropping of elements as easily as the visualization tools. Review collected by and hosted on G2.com.

The platform offers optimized Photon engines on top that perform telemetry processing at great pace for big scale payloads. With serverless SQL warehouses, our teams can run large, concurrent queries without having to manage the scheduling of resources. With the centralized object governance in unity catalog, clear column level data lineage is supported across multi regional workspace deployments. Review collected by and hosted on G2.com.

The web notebook environment doesn't provide native Git branching control and often has state loops which are broken by concurrent code pushes. For complex spatial data visualizations, rendering of such data within active workspace cells will constantly be a memory allocation problem on the browser. Enterprise users identity synchronization configuration menu don't work properly when handling nested group permission rule updates. Review collected by and hosted on G2.com.

Databricks has transformed how our team handles end-to-end data workflows. A few standouts:

UI/UX: The notebook interface is intuitive, and the SQL editor feels polished which switching between Python, SQL, and Scala in the same workspace saves constant context-switching.

Integrations: Native connectors to Azure, Unity Catalog, and Delta Sharing mean we spend less time on plumbing. Lakehouse Federation lets us query external sources without moving data, which was an unexpected win.

Performance: Delta Lake's auto-optimization and liquid clustering noticeably reduced our query times. Photon engine on heavy aggregations is a game-changer for near real-time dashboards.

Pricing/ROI: The DBU model takes getting used to, but consolidating our data warehouse, ETL, and ML tooling into one platform cut our overall infrastructure spend significantly.

Support/Onboarding: Databricks Academy and the built-in documentation made onboarding new engineers faster. The community forum is surprisingly active for niche questions.

AI/Intelligence: Genie (AI/BI) lets business users ask questions in plain English and get accurate results reducing ad hoc requests to our data team by a noticeable margin. Databricks Assistant inside notebooks also accelerates code generation and debugging. Review collected by and hosted on G2.com.

While Databricks is powerful, there are real friction points worth mentioning:

UI/UX: The interface can feel overwhelming for new users have the navigation between Workspaces, Catalogs, and SQL Warehouses isn't always intuitive. Folder and notebook organization could be more structured out of the box.

Integrations: Some third-party connectors still require manual configuration and custom code. Lakehouse Federation is promising but occasionally inconsistent with certain source systems, needing extra troubleshooting.

Performance: Cluster startup times remain a pain point cold starts on interactive clusters can disrupt fast-paced workflows. Serverless compute helps but isn't universally available across all features yet.

Pricing/ROI: The DBU-based pricing model lacks transparency for newer teams. It's easy to rack up unexpected costs without careful cluster policies and monitoring in place. A more straightforward cost estimator would help significantly.

Support/Onboarding: Enterprise support response times can be slow for non-critical tickets. For complex architectural issues, getting to the right expert often takes multiple escalations.

AI/Intelligence: Genie works well for standard queries but struggles with complex multi-table logic or domain-specific terminology without significant fine-tuning. The Databricks Assistant inside notebooks occasionally generates outdated or incorrect API suggestions. Review collected by and hosted on G2.com.

I find the recent opening of the dashboard-building apps invaluable for presenting market findings to our brokerage teams. The Genie feature is great for quick reads across regional pricing datasets without disappointment. Databricks scales effortlessly, which is really helpful when comparing years of transaction records against current inventory. Building a comparative pricing dashboard in much less time than before has been a game-changer for me, with the scanning speed alone reshaping my workflow. Review collected by and hosted on G2.com.

My one sticking point concerns Iceberg interoperability. Our Delta tables read cleanly into Snowflake as Iceberg, yet the reverse path from Snowflake into Databricks has not worked for us. When some of the external datasets I depend on live on the Snowflake side, that gap means I occasionally wait on figures I would rather have at my fingertips, and I wish it had been sorted out sooner. Smoother two-way Iceberg support would help most, so that tables created as Iceberg in Snowflake could be read natively in Databricks the same way Delta tables already flow the other direction, sparing me the wait on external datasets that currently sit out of easy reach. Review collected by and hosted on G2.com.

I like Databricks for its scalability and all-in-one environment for data engineering, analytics, and machine learning. It allows me to process large datasets efficiently while keeping workflows organized in one platform. The scalability is very valuable because it lets me handle growing data volumes and complex workloads without performance issues. As projects expand, the platform can scale resources efficiently. Review collected by and hosted on G2.com.

Some features can have a learning curve, especially for new users working with advanced configurations or cluster management. The interface could also be more intuitive in certain areas. The setup was relatively smooth for core features, but some advanced settings like cluster optimization, permissions, and integrations required more time and technical knowledge. Review collected by and hosted on G2.com.

Databricks provides a unified platform and is very efficient working with large scale terabytes level data. I also like the integration with various cloud services which is seamless and very helpful. Also, the inbuilt Apache spark and very efficient AI/ML workflow orchestration stands out from others. And the databricks support has been outstanding in case of any issues. Review collected by and hosted on G2.com.

With features comes cost and using databricks at a scale we use it (terrabytes data, multi customer, multi environment) becomes cost challenging. Also, learning curve can be bit steep for new beginners. Review collected by and hosted on G2.com.