$500k in Savings and 10x Scalability: Hatchet’s Database Story with Ubicloud

3 min read

Ubicloud eliminated our IOPS bottlenecks and is already saving us $300k a year, with $500k projected as we migrate more workloads.

Alexander Belanger, Founder, Hatchet

Hatchet is a platform for running anything async; workflows, data pipelines, AI agents, and background jobs. Developers integrate Hatchet via its SDK, letting Hatchet’s orchestration layer handle execution outside the web server.Today, Hatchet runs over 100 million tasks per day and stores 15TB of data across its databases. At this scale, database performance and costs became a critical challenge.

Eliminating IOPS Bottlenecks

One of Hatchet’s database shards routinely spiked past AWS RDS limits. To cope, they maxed out at 64,000 provisioned IOPS, but still faced bottlenecks.

“We would keep hitting the IOPS limit on RDS. We didn’t need 64k IOPS all the time, but AWS forced us to stay at that level—and the bill was just way too high.”

Alexander Belanger, Founder, Hatchet

By moving to Ubicloud, Hatchet eliminated IOPS concerns entirely. NVMe-backed storage handled bursts effortlessly, and even during 10x traffic spikes the database continued to perform without alarms or queue build-ups.

Flexible Storage That Adapts to Your Needs

At one point Hatchet needed to scale RDS storage from 1TB to 10TB. However, when the need subsided, AWS offered no easy way to scale down, leaving them stuck paying for capacity they no longer needed.

“It seems like such a simple feature, but once you provision storage in RDS it only goes up. We were paying way too much because we couldn’t scale back down.”

Alexander Belanger, Founder, Hatchet

Ubicloud allows scaling storage down safely and with minimal downtime, giving Hatchet the flexibility to right-size their databases as workloads change and avoiding the waste of overprovisioned capacity.

Massive Cost Savings at Scale

Hatchet’s RDS costs had climbed to around $60k per month, largely from storage, networking, and provisioned IOPS. With Ubicloud, those costs have dropped to the mid-$20k range, cutting hundreds of thousands of dollars per year. This shift translates to over $300k in annual savings, with projections of nearly $500k once the remaining workloads are migrated.

Smooth Migration with Minimal Downtime

Migrations often mean hours of disruption, but Hatchet transitioned to Ubicloud quickly and safely.

“We were able to migrate our core database with just five minutes of paused writes. Having rollback options and support from the Ubicloud team made it stress-free.”

Alexander Belanger, Founder, Hatchet

Using logical replication, the core migration finished with only a few minutes of downtime, while rollback safeguards ensured no risk to production workloads. Ubicloud engineers also actively supported the migration and monitored the process end to end to ensure success.

The Results: More Performance, Lower Spend, and Full Confidence

Hatchet’s migration to Ubicloud produced immediate, quantifiable benefits:

  • No IOPS bottlenecks: NVMe storage handled 10x traffic spikes with smooth, uninterrupted performance.
  • Massive cost reductions: Monthly database spend dropped from $50 - 60k to mid $20k, delivering $300k in annual savings now, with $500k projected as more workloads migrate.
  • Flexible storage management: Capacity can be scaled down safely, avoiding waste and unnecessary expense.
  • Fast, safe migrations: Core databases moved with just 5 - 10 minutes of downtime, backed by rollback safeguards and expert support.

Ready to rethink your cloud?

Discover how Ubicloud can help. If you have any questions, you can reach out to us at [email protected].

SUMMARY

10x
Scale
$500K
Cost reduction
Hatchet reduced its annual database costs by $300k, and expects to save nearly $500k after moving the rest of their databases to Ubicloud. Their infrastructure backed by Ubicloud PostgreSQL now runs 100M+ tasks daily and can absorb 10x traffic spikes thanks to lack of IOPS limits. The final switchover required only a few minutes of downtime. The result: predictable performance, massive cost reductions, and the freedom to scale workloads.
Share this story

Reduce costs. Avoid lock-in.
Win back your infrastructure.

Try it now