Estuary

My main use case for Estuary is to target replacing slow, batch-heavy ETL pipelines with millisecond latency change data capture pipelines for our primary real-time ETL streaming platform for our transactional databases. We use PostgreSQL, and we are using a streaming pipeline Kafka direct with our vector DB, Pinecone, and analytical data warehouse, Snowflake. This allows us to feed fresh transactional data and embeddings correctly into the downstream retrieval-augmented generation RAG, giving us real-time AI agents without writing custom pipeline code or maintaining heavy infrastructure.

The shift from batch-heavy ETL to real-time change data capture has positively impacted our team and workflows, as we have seen a 45% reduction in cloud data pipeline compute cost by migrating from batch-based jobs to continuous CDC. We have saved around five to six weeks of pipeline deployment time for our data engineering and AI teams, which is a great time saver for us. We have seen sub-100ms latency on data availability for our live interface models compared to the legacy hourly batches which used to run earlier, with ultra-low latency that is scalable and no latency delays we have seen so far.

Estuary has positively impacted our organization as we have seen the infrastructure and licensing costs reduced by at least 68 to 72%. The pay-as-you-go billing model has saved us a lot of time, as the engineering teams are not spending weeks writing and maintaining the ETL integration for real-time vector embeddings, which has saved the engineering teams at least 300 hours per year due to the automation it provides.

The changes brought by Estuary have greatly affected my team's productivity and morale, improving our overall project delivery velocity by 28% and decreasing project delivery timelines by 18 to 20%, demonstrating the business impact it has made.

Regarding the best features Estuary offers in my experience, I love that it uses ingesting streaming data via Change Data Capture, which is ideal for our workloads. I also love that it automatically infers and adapts to the schema updates on our source DB without breaking down the downstream consumer or vector indexes. Another great feature is the flexible developer controls, such as CLI commands, SQL commands, live script, and support for Python derivation for transformations. It enforces strict exact processing guarantees and ensures that sensitive data is safe and not leaving our cloud boundary during transit.

Regarding Estuary's AI capabilities, the governance and security of the AI agents and workflows are really strong, making sure that sensitive data remains within the system, with great encryption measures in place to ensure that no bad actor can access it and that the data stays safe within the organization.

Estuary has definitely met my expectations regarding the accuracy and reliability of output, delivering great results while saving a lot of time, and the output is clear and concise.

Estuary integrates really well with the other AWS services we are using, whether it's our databases or internal repositories, connecting easily without any issues in integration.

While I feel Estuary is doing the job very well, if I had to suggest an improvement, I would want to see more detailed dashboards around the pipeline metric visualization in the web console, which is the only enhancement I can think of.

I have been using Estuary for our primary real-time ETL streaming platform for our transactional databases for more than ten years.

Estuary is definitely a stable solution, as we have been using it for more than two years without experiencing any stability issues in any of our environments.

Estuary is definitely a scalable solution; we have over one million users, and it can handle heavy loads, databases, and pipelines efficiently.

Customer support remains one of my favorite aspects of Estuary; when we faced a problem last month regarding the setup of one of our devices, the Estuary team resolved that query within 45 minutes of raising the concern, which is exceptional. I would rate the customer support a ten on a scale of one to ten.

We did not use a different solution prior to Estuary.

The deployment of Estuary in my environment is really easy, with no problems at all; the team helped us to maintain and create a proper deployment into our existing workflows, which integrated easily with our system and the pipelines using MCP connectors. I found the configuration process easy, as we were able to configure the solution within three days of purchase, resulting in a great ROI so far.

We have seen a 45% reduction in cloud data pipeline compute cost by migrating from batch-based jobs to continuous CDC. We have saved around five to six weeks of pipeline deployment time for our data engineering and AI teams. Infrastructure and licensing costs have been reduced by at least 68 to 72%. The engineering teams have saved at least 300 hours per year due to the automation provided. Our overall project delivery velocity has improved by 28% and project delivery timelines have decreased by 18 to 20%.

The procurement process was really easy, with no problems at all. The metering and billing experience is the best thing we have seen, as we opted for the pay-as-you-go model, which is saving us a lot of money and time.

Before choosing Estuary, we evaluated other options such as Stream, Confluent Cloud, and Fivetran.

I rate Estuary a five on a scale of one to five based on the experience we have had, the ROIs that were delivered, and the features it contains, which justifies my rating. My advice to others looking into using Estuary is that anyone seeking a solution that can automate their pipelines should definitely consider Estuary. I give this product an overall rating of ten.

Amazon Web Services (AWS)

What do you like best about the product?

The support from Estuary was excellent. There was an issue with the first connector, which was resolved within a few hours together with a developer.

Communication in Estuary's Slack channels works excellently. You can use AI for simple questions, and if you can't get any further, support responds quickly. You just have to keep in mind that Estuary operates in North America and its time zones, which is currently a disadvantage for a European company.

The user interface is very simple and easy to understand.

What do you dislike about the product?

Perhaps the only thing Estuary currently lacks is more connectors to large and established ERP systems. Integration into enterprise environments is also still expandable. The idea of a dedicated VPN and automatic management is good and right, but there is still a gap between enterprise and "Estuary Standard"; I still miss an offer in between.

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

We use Estuary to reliably transfer data from our ERP, CRM, and internal systems to Snowflake.

What do you like best about the product?

Low-maintenance, solid availability for syncing data from disparate data sources.

What do you dislike about the product?

Some limited observability - would be nice to see more detailed logs, and have more insight/direct queryability into the individual data collections. Also the BYOC model is a bit annoying, would be nice to be able to use our existing cloud VPC.

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

Our company has a LOT of different data stored in very different places, so this is helping us get it all into one.

What do you like best about the product?

Switching our data pipelines from Fivetran to Estuary was one of the best infrastructure decisions we’ve made. It drastically lowered our monthly bill while also giving us true real-time streaming capabilities. Migrating to Estuary immediately stabilized our spend with a far more predictable, volume-based model.

Moving away from delayed batch syncs to continuous, low-latency streaming completely upgraded our data freshness, letting operational data flow virtually instantly. Estuary also gives us the flexibility to customize schemas and handle in-flight transformations directly within the pipeline, which has helped us save on downstream compute costs.

While the initial setup and schema mapping required deliberate effort, day-to-day management has been remarkably smooth and genuinely “set-and-forget.” Ultimately, Estuary delivered the blend of real-time performance, developer control, and budget relief we needed as we scaled.

What do you dislike about the product?

What I don’t like about Estuary is the UI. At times it isn’t very intuitive. For instance, when I’m backfilling records, the progress details aren’t very clear and are mostly shown as a percentage, which doesn’t always help me understand what’s actually happening. The logs are also pretty noisy and hard to follow, especially when I’m troubleshooting. Since I’ve had to work through capture failures, binlog position changes, and backfills during our migration, clearer progress indicators and more structured, easier-to-scan logs would make the overall experience much better.

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

Estuary has been highly effective in supporting our cost-reduction strategy while still delivering a robust, reliable, high-quality data ingestion service. Migrating from Fivetran to Estuary helped us significantly reduce our data pipeline costs without sacrificing reliability or performance.

Data ingestion and loading have been largely seamless, and the platform has handled our production workloads well. We’ve also found the Slack support team responsive and helpful, especially when troubleshooting issues related to captures, backfills, binlog positions, and migrations. The AI-powered support through Kapa.ai has been useful as well, letting us quickly find answers and navigate the documentation.

Overall, Estuary strikes a strong balance between cost efficiency, reliability, performance, and developer support, and it has become a valuable part of our modern data stack.

What do you like best about the product?

I like how Estuary is a key element within our data stack, making data ingestion from all of our Postgres data sources and Salesforce reliable. It's easy to use, and I appreciate the strong communication when something is broken. Their support is very quick to help us, and it's nice knowing there's a part of our stack we don't have to worry about. Estuary just works, allowing us to focus on other high-priority tasks. I also found the initial setup of Estuary to be very easy. Support is readily available if we have questions, which is really helpful. The reliability of Estuary makes everything a lot easier.

What do you dislike about the product?

Sometimes I find it hard when we want to add a new data source or a connector because it's like $200 per connector, which has made us purposely not use Estuary at times. I wish there was a way to onboard a source with pricing charged per active row instead, so we could use it without having to think about the cost.

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

Estuary reliably ingests data from Postgres and Salesforce, simplifying our workflow. Its ease of use and strong support mean we focus on priorities without worry. It's a key part of our data stack, making us trust that it just works, saving time and effort.