PhoenixAI (formerly CelerData)

The Agentic AI Database. Giving autonomous AI agents sub-second access to live enterprise data at massive scale. Built for production. Formerly CelerData.

[PhoenixAI (formerly CelerData) @CelerData]

Still thinking about our PhoenixAI happy hour at Ula 🐦‍🔥

Thank you to everyone who came out for great data + AI conversations, new connections, and Phoenix Margaritas made just for the event 🍹

Congrats again to our AirPods raffle winners 🎧

See you at the next one!

[PhoenixAI (formerly CelerData) @CelerData]

🐦‍🔥 Meet Agent Fawkes 👉 https://t.co/j6PYLEGJUJ — the AI copilot inside PhoenixAI Cloud.

Ask in plain English → runnable SQL. Fix broken queries in one click. Optimize, explain, and chart results without leaving the editor. And the model never sees your rows (BYOC).

[PhoenixAI (formerly CelerData) @CelerData]

Last call, and we mean it 🥂

In SF tomorrow? Skip the big networking thing for an hour and come to Ula instead, 6–8pm.

Small room, good drinks, nobody scanning a badge.

We're approving RSVPs one by one and seats are almost gone. Sneak in here👇

https://t.co/NSl8Y3VBm9

[PhoenixAI (formerly CelerData) @CelerData]

Some months you ship a feature. This month we shipped a whole new identity. 🐦‍🔥

The May issue covers the rebrand, what's new in PhoenixAI Cloud (incl. Agent Fawkes), the latest customer stories in prod, and June 11 launch webinar.

The full recap 👇

https://t.co/qUmEgByb86

[PhoenixAI (formerly CelerData) @CelerData]

You've heard the name. Now come see what PhoenixAI is all about. 🐦‍🔥

June 11: live session with Sida Shen — what PhoenixAI is, why we're building it, and how it supports the full agent loop (decision → execution → trace).

Bring your questions 👇

https://t.co/FPWwgzwR1N

[PhoenixAI (formerly CelerData) @CelerData]

LLM handles the conversation. PhoenixAI handles the truth.

How Conductor built a 100%-decimal-accurate #AI agent on terabytes:

▸ Sub-second queries (389ms–2.8s)
▸ Idea → prod in <2 months
▸ Split-reasoning, deterministic every run

Full Story👇

https://t.co/wAcnhF4Sm3

[PhoenixAI (formerly CelerData) @CelerData]

Real-time ads. Faster BI. An AI agent that actually works.

CTW (the team behind G123) replaced TiDB + direct lakehouse access with PhoenixAI:

✅ <10s audience freshness
✅ 3× faster Tableau on Delta Lake
✅ 90%+ AI agent accuracy

👇
https://t.co/Sg7WFokisc

[PhoenixAI (formerly CelerData) @CelerData]

PhoenixAI serves AI agents the way they actually query: ad-hoc, sub-second, at 10K+ concurrent QPS on live enterprise data.

The #RealTime Analytical Database for AI agents. In your cloud, or self-hosted.

👉 https://t.co/26wdGQSnBX

[PhoenixAI (formerly CelerData) @CelerData]

CelerData is now PhoenixAI. 🔥

Same team, same database, same commitment — sharper focus on what we've always been building toward: the analytical database for the age of AI agents.

Welcome to PhoenixAI 👉 https://t.co/9b8yUJzABy

[PhoenixAI (formerly CelerData) @CelerData]

⚡Every second counts for 100M+ customers. Intuit IPS delivers 100k+ events/sec with TP99 <500 ms and <4 s freshness. ML ~2 s e2e. Druid → StarRocks = simpler ops, fast joins, native upserts, 98.33% faster aggs.

Get the full scoop👇 https://t.co/qpoG4RCzyt

[PhoenixAI (formerly CelerData) @CelerData]

AI Council SF 2026 is tomorrow — May 12–14 in San Francisco. 🎟️ RSVP: https://t.co/pjIbdRq6Qs

We’re looking forward to conversations with the teams actually building AI and data infrastructure — what’s working, what broke in prod, and what’s coming next！

[PhoenixAI (formerly CelerData) @CelerData]

Less than a week to go. We're heading to AI Council SF 2026 — May 12–14 in San Francisco. 🎟️ RSVP: https://t.co/XZj7vH5EQj

Connect with the people actually building AI and data infrastructure — and hear what's working, what broke in prod, and what's coming next!

[PhoenixAI (formerly CelerData) @CelerData]

Next Thursday, May 14 — we're joining @confluentinc in Chicago for the Data Streaming World Tour

🎟️ Free to attend — grab your spot: https://t.co/PkXag0qMUD

[PhoenixAI (formerly CelerData) @CelerData]

🌟 40s → 800ms. 480K API calls → 80K. 8 weeks of engineering time back every year.

Those are Demandbase's numbers after migrating 400M+ records from Elasticsearch to CelerData — and they shared exactly how they did it.

Read the full story 👇

https://t.co/BWHSPrbA34

[PhoenixAI (formerly CelerData) @CelerData]

✅ ~90% analytics cost reduction
✅ Up to 95% Snowflake usage reduction
✅ Sub-second dashboards

Fanatics simplified analytics into one open lakehouse: Iceberg + StarRocks. The blog walks through the “why” and “how.”👇

https://t.co/FPoSGlR6qy

[PhoenixAI (formerly CelerData) @CelerData]

📬 The April issue is out! New CelerData Cloud BYOC features, a StarRocks 4.1 release you'll want to dig into, customer stories that pull back the curtain, community work pushing the ecosystem forward — plus where you can find us in person next month 👇

https://t.co/fzlKwtKiCZ

[PhoenixAI (formerly CelerData) @CelerData]

As AI agents take center stage, traditional data architectures are being pushed to their limits. Explore the core challenges of agent-driven analytics—and the forward-thinking approaches designed to meet them—in our blog 👇

https://t.co/8BRm93DIbL

[PhoenixAI (formerly CelerData) @CelerData]

Customer-facing analytics is one of the hardest workloads in data infra. Sub-second latency, multi-table joins, real-time updates, multi-tenant skew — all while users tap impatiently inside your product.

A breakdown of how to pick the right engine 👉 🎥 https://t.co/GIEK1qaswk

[PhoenixAI (formerly CelerData) @CelerData]

Discover how Pinterest leveled up their analytics by migrating to StarRocks, cutting p90 latency by 50% and using just 32% of their previous infrastructure. They saw a 3x improvement in cost-performance and now enjoy data freshness in just 10 seconds.🚀

https://t.co/l8OonjThZ5

[PhoenixAI (formerly CelerData) @CelerData]

👀 Crypto data isn’t just “big” — it’s open, noisy, and mission-critical for compliance. Coinbase rebuilt their analytics stack to trace money flows across 10+ blockchains, run joins on 573B+ rows, and keep fraud detection real-time.

Full story👇
https://t.co/neJ8bNesqN