Alexis Rico 🦋

Principal Software Engineer @xata 🦋 | Also building ⚡

[Alexis Rico 🦋 @SferaDev]

Been working on PgBeam, a PostgreSQL proxy that pools connections and caches queries across multiple regions.

Any driver, any ORM, just swap the connection string. Live benchmark running real queries from 20 regions: https://t.co/NYDcQrXQjb

[Xata 🦋 @xata]

Introducing the latest in Postgres for agentic workloads.

Agents can now branch or wake your Postgres database instantly.

[Alexis Rico 🦋 @SferaDev]

@MatthewStanciu @vercel Exciting!

[Tudor Golubenco @tudor_g]

We needed a storage layer that can scale to a huge number of volumes (think millions) for our Postgres platform.

Most of the existing storage systems are optimized for a few volumes with really high performance. But we needed the opposite: a very large number of mostly idle volumes.

So we wrote our own.

It’s called Xatastor and it enables Postgres-per-tenant use cases, “ephemeral” dbs for agents, free tiers, etc. Supports copy-on-write snapshots, clones, and thin provisioning.

It’s based on ZFS and NVMe-oF as key technologies. Link with all the details in the first reply.

[Alexis Rico 🦋 @SferaDev]

The future of databases in the post-AI age. At @xata we believe in this vision and can't wait to share more on how Copy on Write unlocks agentic workflows with production-like data at minimal cost.

[Netlify @Netlify]

Meet the new Netlify Database.

Fully managed Postgres, built into the platform. Every agent run automatically gets its own database branch.
Here's why that matters. 🧵

https://t.co/zrtVgZlCPT

[Alexis Rico 🦋 @SferaDev]

@AvgDatabaseCEO

[Xata 🦋 @xata]

Xata is now open source. Apache 2.0.

Postgres platform with Copy-on-write branching at the storage layer. Copy a TB-sized database in seconds. Inactive copies scale to zero automatically. 100% vanilla Postgres.

Two use cases driving this:

• Preview and testing environments with real production data
• Platforms provisioning per-user Postgres at scale

Built on CloudNativePG + OpenEBS. Runs on Kubernetes.

What's in the release:

• SQL gateway, responsible for routing, IP filtering, waking up scaled-to-zero clusters, serving the serverless driver over HTTP / websockets, etc.
• Branch operator managing all resources related to a branch.
• Clusters and projects services for the control-plane and REST APIs
• Auth service, based on Keycloack for API keys
• CLI that makes use of the REST API
• Scale-to-zero CNPG plugin for automatically hibernating branches on inactivity

This is the first of several announcements. More next week.

[Alexis Rico 🦋 @SferaDev]

@cramforce @zahlekhan @rauchg Going to sleep now but tomorrow I'll revisit the benchmark approach so that execution is no longer server-to-server in Vercel Functions and that the origin request does not even come from the AWS network. That should give results that are the closest to reality.

[Alexis Rico 🦋 @SferaDev]

"Deploy globally" is a lie when your database is in us-east-1.

The next big infrastructure unlock isn't faster compute. It's making the data layer invisible.

Curious if @rauchg and @cramforce see this the same way.

https://t.co/YmYyGz0qgw

[Alexis Rico 🦋 @SferaDev]

@cramforce @zahlekhan @rauchg Did a quick experiment trying to add non-trivial writes to the benchmark and also comparing to compute at origin.

Round-trip times are comparable and in some cases better. It's all server to server, so IRL latency difference would be even bigger.

https://t.co/fdZHDNHz8U

[Alexis Rico 🦋 @SferaDev]

@zahlekhan @rauchg @cramforce For write operations you benefit from TLS resolution near you and avoiding multiple RTTs, which is quite a lot already.

Also PgBeam handles origin to source region data transfer in its own network that is faster than if you connected over regular internet.

[Alexis Rico 🦋 @SferaDev]

@cramforce @rauchg Appreciate it! It's a problem I keep seeing working for a database company, so I started building it as a side project.

Hoping to make it sustainable enough and scale it to be production ready, early numbers are encouraging.

Open to chat if you ever want to discuss ideas!

[Alexis Rico 🦋 @SferaDev]

@connyb @ctatedev Nope, didn't see that one! Feel free to open a PR and contribute it as an official adapter!

[Alexis Rico 🦋 @SferaDev]

@vercel_dev KYC!

[Alexis Rico 🦋 @SferaDev]

@jeffwilcox Diversification at its best 😅

[Alexis Rico 🦋 @SferaDev]

@micLivs @ctatedev https://t.co/0lgfZ3PKXM

[Alexis Rico 🦋 @SferaDev]

Weekend project: terminal renderer for @vercel json-render using ink.

Same JSON spec that drives web UIs now renders fully interactive terminal apps: text inputs, selects, markdown, tables, and more.

The spec patches in real-time as the model generates, so you can see components stream progressively instead of waiting for the full response from the LLM.

[Alexis Rico 🦋 @SferaDev]

@Utkarsh51557661 @ctatedev At @xata we plan to use it in our CLI as a way for the LLM to return questions about the database similarly how you would do it in the web UI. So that the CLI is just another entrypoint to the same agentic experience.

[Alexis Rico 🦋 @SferaDev]

@RicheTechGuy @ctatedev On it!

[Alexis Rico 🦋 @SferaDev]

@pelaseyed @ctatedev I'll take a look into OpenTUI as well and add support to it. JSON Render is a framework agnostic library so one doesn't imply you don't have the other!

[Alexis Rico 🦋 @SferaDev]

Weekend idea shipped to upstream open source project in less than 24 hours, we're cooking!

Thanks @ctatedev for the awesome @vercel labs projects! This is just the beginning, let's see where we can take this 🫡

[Chris Tate @ctatedev]

Introducing Generative TUI

Ask anything - get polished dashboards with real data, rendered live in your terminal.

27 components. Streaming. json-render + Ink.

npx skills add vercel-labs/json-render --skill ink