Presteniko Septi R

I gave up taking all of the photos lol but I really enjoy what @gappy50 さん was talking about. Especially about the usage of the cortex code cli. I learnt a lot. Thank you. 👍

Yes our team's ultimate goal is to improve the data management lifecycle and workflows.

@gappy50 さん's presentation is starting now 👍👍👍

今度は今度。今は今。

いつもニコニコNiko (@nikoraha) さんの記事💣 ワークフローと定型業務の分類ができれば散らばったコンテキストもrepoもskillsで一撃です💅 データ基盤チケットを自動化した話: Notionからデプロイまで、Claude Code Agentで一気通貫｜Presteniko Septi Rahadian zenn.dev/dely_jp/articl… #zenn

The more I read about LLM engineering and its implementations, the more I believe this field of engineering is so interesting. KV cache? Scaling contexts in each inference? Now that's something that I really want to touch.

Can't wait to play with Gemma 4.

My Nakama wrote this blog. It is such a useful way to actually measure performance in Snowflake. Really cool!

Just to count my blessings I’d actually need an accountant. It's not all about the money; more than that.

Okay, let's think outside of the current hype and claims that AI could insanely increase productivity. I'm curious how we can seriously measure productivity gains with AI? What productivity that can or can't be measured?

In Claude Code, can we put files as our "claude.md" in artifacts to save memories and contexts? Interesting.

NEW AI report from Google. Every prior intelligence explosion in human history was social, not individual. These authors make the case that the AI "singularity" framed as a single superintelligent mind bootstrapping to godlike intelligence is fundamentally wrong. This is directly relevant to anyone designing multi-agent systems. They observe that frontier reasoning models like DeepSeek-R1 spontaneously develop internal "societies of thought," multi-agent debates among cognitive perspectives, through RL alone. The path forward is human-AI configurations and agent institutions, not bigger monolithic oracles. This reframes AI scaling strategy from "build bigger models" to "compose richer social systems." It argues governance of AI agents should follow institutional design principles, checks and balances, role protocols, rather than individual alignment. Paper: arxiv.org/abs/2603.20639 Learn to build effective AI agents in our academy: academy.dair.ai

Putting a harness and guardrail is beneficial in data modeling development using AI, but for the critical models, imposing data contracts on evals would be a compelling idea.

AIを中枢として会社をどう作るか？という問いを考え、実装し続けています。クラシルではデータチームが先駆けて「AI Readyデータ基盤」を構築することに力を入れています。 ・Tier定義で品質を段階管理（Tier3以上＝AI利用可能） ・Claude Code Agent Teamで設計・実装・検証を自律化 ・人間は要所で判断し、チーム自体を育て続ける仕組み これにより、少人数でも「品質×スピード」を実現しています。AIエージェントが本当に価値を発揮できる土台を、エンジニアリングで固めています。 AI時代に求められるエンジニア組織、データ基盤を逆算して作っていきます。

Streamlitに人生をかけているfunzinさんの活用術。 Snowflakeとの相性の良さもありますが、Claude Codeからのデータ活用は、AI活用を前提としたデータ管理が正しく行えているかだと思うので、データチームも引き続き頑張ります。

私のMentor、@gappy50 さんのコメントを読んでみてください！

Your semantic layer just got better 👀 It can now fix itself - spotting broken metrics, suggesting corrections, and improving with every query. Stop babysitting dashboards and let your BI tool do the work. Learn more 👇 lightdash.com/blogpost/your-…

🚀 Big news for data teams: Lightdash now supports @ClickHouseDB! Because a fast warehouse deserves a fast BI tool 😏 Learn more about the integration here 👉 #clickhouse?utm_source=x&utm_medium=social&utm_campaign=clickhouse" target="_blank" rel="nofollow noopener">docs.lightdash.com/get-started/se…

Mood