In my opinion, Snowflake Analytics can be improved by introducing more features, such as additional integration options. I remember using Snowflake Pro, which allows exporting direct data into the S3 bucket, but there were a couple of things that were lacking at that time, particularly data transformation during exports, and I would suggest adding some minor features. I would like to see more integration features, specifically with Amazon Web Services, in the next release of Snowflake Analytics. Most companies are using these cloud technologies, so additional features and some transformation capabilities while exporting to different data buckets would be very helpful.
There are some minor issues encountered with Snowflake Analytics, such as problems when working with identity or row number generating different results or issues with referential integrity, which are being reported on the Snowflake Analytics website, indicating they are looking into these problems for future updates. I would prefer Snowflake Analytics to improve their support response times, as sometimes the responses we receive are not very prompt and ticket assignments may not be timely.
The advantages of Snowflake Analytics outweigh the disadvantages. However, if it offered flexibility similar to Oracle and supported more heterogeneous data sources and database connectivity, it would be even better. PySpark is emerging as a competitor, so Snowflake Analytics needs to bridge any gaps between them.
Most features are already available. Perhaps AIML integration might be further needed, or Snowflake is already implementing it. AIML-based SQL prompt and query generation could be an area for enhancement.
RSA Country Manager at a non-profit with 1-10 employees
Reseller
Top 5
Sep 3, 2024
Snowflake Analytics should probably have more built-in tools for master data management, data purification, and enhancement. Instead of using third-party tools, it would probably be good if it could be internal to Snowflake. So, to have integrated capabilities for master data management.
I don't see many drawbacks with Snowflake Analytics, but it's not as mature as other tools. It is evolving and needs to integrate various features, like data loading and analytics, better. These components are not fully connected, so the tool should become a more integrated application. The solution is costlier than Azure or other tools like Pentaho, which are cheaper but require more effort in ETL or ELT development and integrating all the tools.
In terms of cost. Many customers face issues with the expenses on Snowflake. The pricing visibility is complex. If you understand pricing, you can estimate costs, but if not, it can be challenging. They need to provide better cost visibility upfront. For example, credits are not always utilized properly, which isn't easy for new users. A more user-friendly cost calculator would help, as the current one requires expertise to use effectively. These things are not easy for someone who is new to Snowflake, unlike Azure's calculators, which give good visibility into the cost. It does have a calculator, but you need to be skilled in using it. Otherwise, it's hard to estimate the total cost for future loads. For additional features, there could be its own AI model like Google has Gemini. Snowflake can come up with those kinds of features for Analytics. If you can just type in your prompt and get answers, especially for Analytics people on the client side, they will love that. Like, in Google's Gemini, you type the prompt and get the answers based on the data available. If they come up with that kind of AI and UI, it will be easier for clients to analyze their data. They can just type a query like, "What is my highest selling product?" They don't have to write a complex query.
I haven't noticed any limitations with the solution. There could be more analytics. We find that IBM has a lot of pro data analytics that we use. The distribution methodology isn't as strong as Bethesda or SAP HANA. It's not as strong as other competitors.
Technical Consultant at a financial services firm with 10,001+ employees
Consultant
Top 20
Apr 9, 2024
Some functionalities available through SQL are inaccessible in Python or Java within Snowflake. Snowflake must enhance its support for these programming languages to offer full flexibility and capacity, allowing developers to use them. Currently, there are restrictions. For example, using Spark Engine with Snowflake allows you to write code but not call external functions or packages. Snowflake must improve its programmatic capabilities to support better engineers and developers, especially those transitioning from big data technologies.
Associate Director - Delivery (Technology DWH & Data Engineer) at MOBIUS KNOWLEDGE SERVICES PRIVATE LIMITED
Real User
Apr 5, 2024
Implementing everything on-premise is challenging because it require proper support from advisors, DBAs, and others. However, with cloud implementation, there are fewer technical barriers.
The scheduling of jobs requires improvement, particularly in terms of the user interface which currently lacks certain features found in comparable platforms.
Snowflake's Snowpark is an area of concern where improvements are required. Snowflake's Snowpark is a pretty new concept. Snowflake's Snowpark should improve my adding new libraries and getting the codes deployed quickly.
Technical Lead at a computer software company with 10,001+ employees
Real User
Feb 13, 2024
Snowflake could improve in the areas of advanced machine learning AML and generative AI. While it excels in data warehousing, it lags in these areas compared to other platforms. Snowflake needs to evolve further to gain more traction in the data science space.
Machine learning should be improved. Some of the solutions, such as Databricks, offer an option to directly write and integrate Python code on the spot, but Snowflake doesn't have that feature yet. Adding it would be useful, as it would facilitate building models.
Assistant Manager Operations at Tech Mahindra Limited
Real User
Jul 12, 2023
We are satisfied with the solution since Snowflake brings in a lot of new features now and then. The solution's high price can be an area of concern that needs improvement.
The tool should support EIM use cases. I guess the product is already working on it. I look forward to seeing inbuilt AI generative tools in the solution's future releases. The tool's price can be a little lower. The solution's on-premises support is also very limited. We have to rely on other support services to deploy it on-premises.
Snowflake should include a WHERE clause for building data pipelines. When we create pipelines, Snowflake does not provide the WHERE clause. So the data transformation does not take place. We have to dump data as it is and transform it into the pipeline. If the WHERE clause is included, it will be very beneficial in the future.
There are issues while loading data from Snowflake Analytics to the Power BI reporting, and it doesn't get loaded directly. Sometimes we need to use queries to fix it, but it affects our reporting. Snowflake Analytics should be easily loaded into the reporting tool because it is the final end tool.
For modeling of data as well as for AI and predictive analytics, they need to improve their integration into different Python and Jupyter notebooks. That is lacking but I understand they're working on it.
Snowflake Analytics can improve the integration with machine learning tools and AI and it will make the solution more usable. In a future release of Snowflake Analytics, transactional processing should be supported natively within the solution. The transactional process means, for example, you go to a bank, you draw money, you deposit money. These are all real-time transactions. Thousands or millions of people who draw money and deposit money participate in using transactional systems, such as ATM withdrawals. We want this to be supported by the Snowflake system.
End-to-end execution of jobs isn't possible with Snowflake, which means we have to do some customization. Allowing sequential flow of orchestration would be an improvement. In the next release, Snowflake should include direct integration and communication between other cloud systems.
Snowflake Analytics offers advanced capabilities in data warehousing and cloud data migration, with support for machine learning and business intelligence tasks. Its scalable architecture supports large data volumes while enhancing cost efficiency through decoupled computation and storage.As a flexible, managed environment, Snowflake Analytics enhances data sharing and integration across multiple cloud platforms. It allows seamless data pipeline creation, supports advanced analytics, and...
In my opinion, Snowflake Analytics can be improved by introducing more features, such as additional integration options. I remember using Snowflake Pro, which allows exporting direct data into the S3 bucket, but there were a couple of things that were lacking at that time, particularly data transformation during exports, and I would suggest adding some minor features. I would like to see more integration features, specifically with Amazon Web Services, in the next release of Snowflake Analytics. Most companies are using these cloud technologies, so additional features and some transformation capabilities while exporting to different data buckets would be very helpful.
There are some minor issues encountered with Snowflake Analytics, such as problems when working with identity or row number generating different results or issues with referential integrity, which are being reported on the Snowflake Analytics website, indicating they are looking into these problems for future updates. I would prefer Snowflake Analytics to improve their support response times, as sometimes the responses we receive are not very prompt and ticket assignments may not be timely.
I do not see any areas that could be improved with Snowflake Analytics.
The advantages of Snowflake Analytics outweigh the disadvantages. However, if it offered flexibility similar to Oracle and supported more heterogeneous data sources and database connectivity, it would be even better. PySpark is emerging as a competitor, so Snowflake Analytics needs to bridge any gaps between them.
Most features are already available. Perhaps AIML integration might be further needed, or Snowflake is already implementing it. AIML-based SQL prompt and query generation could be an area for enhancement.
When we insert data into a large table, it takes a lot of time. This performance issue can be improved.
Snowflake Analytics should probably have more built-in tools for master data management, data purification, and enhancement. Instead of using third-party tools, it would probably be good if it could be internal to Snowflake. So, to have integrated capabilities for master data management.
I don't see many drawbacks with Snowflake Analytics, but it's not as mature as other tools. It is evolving and needs to integrate various features, like data loading and analytics, better. These components are not fully connected, so the tool should become a more integrated application. The solution is costlier than Azure or other tools like Pentaho, which are cheaper but require more effort in ETL or ELT development and integrating all the tools.
In terms of cost. Many customers face issues with the expenses on Snowflake. The pricing visibility is complex. If you understand pricing, you can estimate costs, but if not, it can be challenging. They need to provide better cost visibility upfront. For example, credits are not always utilized properly, which isn't easy for new users. A more user-friendly cost calculator would help, as the current one requires expertise to use effectively. These things are not easy for someone who is new to Snowflake, unlike Azure's calculators, which give good visibility into the cost. It does have a calculator, but you need to be skilled in using it. Otherwise, it's hard to estimate the total cost for future loads. For additional features, there could be its own AI model like Google has Gemini. Snowflake can come up with those kinds of features for Analytics. If you can just type in your prompt and get answers, especially for Analytics people on the client side, they will love that. Like, in Google's Gemini, you type the prompt and get the answers based on the data available. If they come up with that kind of AI and UI, it will be easier for clients to analyze their data. They can just type a query like, "What is my highest selling product?" They don't have to write a complex query.
I haven't noticed any limitations with the solution. There could be more analytics. We find that IBM has a lot of pro data analytics that we use. The distribution methodology isn't as strong as Bethesda or SAP HANA. It's not as strong as other competitors.
Some functionalities available through SQL are inaccessible in Python or Java within Snowflake. Snowflake must enhance its support for these programming languages to offer full flexibility and capacity, allowing developers to use them. Currently, there are restrictions. For example, using Spark Engine with Snowflake allows you to write code but not call external functions or packages. Snowflake must improve its programmatic capabilities to support better engineers and developers, especially those transitioning from big data technologies.
Implementing everything on-premise is challenging because it require proper support from advisors, DBAs, and others. However, with cloud implementation, there are fewer technical barriers.
The scheduling of jobs requires improvement, particularly in terms of the user interface which currently lacks certain features found in comparable platforms.
Snowflake's Snowpark is an area of concern where improvements are required. Snowflake's Snowpark is a pretty new concept. Snowflake's Snowpark should improve my adding new libraries and getting the codes deployed quickly.
Snowflake could improve in the areas of advanced machine learning AML and generative AI. While it excels in data warehousing, it lags in these areas compared to other platforms. Snowflake needs to evolve further to gain more traction in the data science space.
One notable absence in Snowflake's offerings is an on-premises solution.
The UI could be more user-friendly.
The tool must provide better machine-learning functionalities.
Machine learning should be improved. Some of the solutions, such as Databricks, offer an option to directly write and integrate Python code on the spot, but Snowflake doesn't have that feature yet. Adding it would be useful, as it would facilitate building models.
The platform could work easier for AI implementation compared to Databricks.
We are satisfied with the solution since Snowflake brings in a lot of new features now and then. The solution's high price can be an area of concern that needs improvement.
The tool should support EIM use cases. I guess the product is already working on it. I look forward to seeing inbuilt AI generative tools in the solution's future releases. The tool's price can be a little lower. The solution's on-premises support is also very limited. We have to rely on other support services to deploy it on-premises.
Snowflake should include a WHERE clause for building data pipelines. When we create pipelines, Snowflake does not provide the WHERE clause. So the data transformation does not take place. We have to dump data as it is and transform it into the pipeline. If the WHERE clause is included, it will be very beneficial in the future.
There are issues while loading data from Snowflake Analytics to the Power BI reporting, and it doesn't get loaded directly. Sometimes we need to use queries to fix it, but it affects our reporting. Snowflake Analytics should be easily loaded into the reporting tool because it is the final end tool.
For modeling of data as well as for AI and predictive analytics, they need to improve their integration into different Python and Jupyter notebooks. That is lacking but I understand they're working on it.
Snowflake Analytics can improve the integration with machine learning tools and AI and it will make the solution more usable. In a future release of Snowflake Analytics, transactional processing should be supported natively within the solution. The transactional process means, for example, you go to a bank, you draw money, you deposit money. These are all real-time transactions. Thousands or millions of people who draw money and deposit money participate in using transactional systems, such as ATM withdrawals. We want this to be supported by the Snowflake system.
We haven't seen any areas that are lacking. I can't recall coming across missing features. Over the last two to three years, we haven't had issues.
End-to-end execution of jobs isn't possible with Snowflake, which means we have to do some customization. Allowing sequential flow of orchestration would be an improvement. In the next release, Snowflake should include direct integration and communication between other cloud systems.