Chinh (lelouvincx) lelouvincx

Performance of Showbie is what I’m concerning. Should raise this first to the team and ask for approach.

• Either we should be prepared for blaming question in worst case like why Holistics runs so slow?
• Raised this with anh Huy: https://holistics.slack.com/archives/C08EN0G719N/p1774609569429499
• We just wait for the next call and ask them, then spend time investigating later.

Tried Zed IDE, not supporting ampcode, and I’m more familiar with terminal-based IDE like neovim. Not suitable for me.

A long day with a lot of tasks. Still productive!

Tasks

Partly today

Done

• DONE Showbie - embedding question

• Update document: to add FAQ about why row-level permissions are bypassed if you are admin.

DONE Call with Adam from superb

• To prepare: a demo4 local react app (backend + frontend), similar to what I demo to [[Basata]]
• Note down some feedbacks for embed team:
  • User Attribute:
    • The UI/flow for User Attributes is counterintuitive.
    • Why do internal attribute configurations impact the embedded environment?
    • Ownership Misconception: Adam assumed User Attributes were tied to his personal account/dataset ownership. He questioned if he needed to change the “Dataset Owner” to make attributes function correctly for other users.
  • Embedded Analytics page is difficult to find within the top header navigation.
  • Permission-as-Code: Since permissions are defined via code, there was significant friction regarding where to place the code, the specific syntax required, and what values to input, what do they mean, presentation in List tab.

((69c2a77c-8eb1-4a4b-b87d-ea26d378c713))

Tasks

Partly today

• DONE Share the document for the embedded dashboards for different customers

  • My typical workflow:
    • Tell gemini to design a basic UI based on customer’s branding guide.
    • Tell ampcode implements it, using basic backend/frontend architecture: App.jsx (react) for frontend and server.js (express) for backend.
    • Holistics sets frame-ancestors ‘self’ https: on its portal, which blocks iframes loaded from http://localhost because it only allows HTTPS origins. The browser refused to render the Holistics iframe with the error: “Framing ‘https://us.holistics.io/' violates the following Content Security Policy directive”.
    • The solution was to serve the local dev server over HTTPS using @vitejs/plugin-basic-ssl, so https://localhost:5173 satisfies the CSP policy.
  • How to I present how to set up embed portals to customers?
    • Explain high-level mechanism logseq.order-list-type:: number
      • https://media.holistics.io/347e2a6d-embed-portal-getting-started.png logseq.order-list-type:: number
    • In Holistics, show: logseq.order-list-type:: number
      • Embed portal in development and how to add dataset/dashboard objects. logseq.order-list-type:: number
        • Firstly just develop model/dataset/dashboard as usual. logseq.order-list-type:: number
        • Then put whichever object they want to externally embed into their app. logseq.order-list-type:: number
      • Credentials location. logseq.order-list-type:: number
    • Backend side: logseq.order-list-type:: number
      • Step 1, generate payload, which is basically a JSON object to configure how and which each customer will see. logseq.order-list-type:: number
      • Present parameter reference. logseq.order-list-type:: number
      • Sign payload with JWT token using the secret. logseq.order-list-type:: number
    • Frontend side: logseq.order-list-type:: number
      • Render iframe based on the signed JWT token. logseq.order-list-type:: number
  • Then ask the question: how to restrict data access for each customer? Answer: we come to row-level permission
    • High-level mechanism logseq.order-list-type:: number
      • https://media.holistics.io/9f8f4b64-embed-portal-data-permissions-updated.png
    • Add user attributes. logseq.order-list-type:: number
    • Add permission as code into dataset. logseq.order-list-type:: number
      • Explain the single entry point for row-level permission is dataset. Once add into dataset, every widgets of a dashboard query from that dataset will be restricted by RLP. logseq.order-list-type:: number
      • If a dashboard is not restricted, it’s because that dashboard is using another dataset. logseq.order-list-type:: number
    • See preview in development embed portal. logseq.order-list-type:: number
    • Configure payload to include that user attribute. logseq.order-list-type:: number

Done

• DONE Ask squad modeling why the user 112390 does not exist

IDEA synthesize tone and writing style of the everyone in Holistics (resource: notion) and extract key insights from it, then materialize into an agent skill for better notion skill’s output

Tasks

Partly today

• ((69bcb599-a91e-49ea-bc3d-6a5a651f03df))
• ((69c2a77c-2f5b-4c6e-851b-4d8425fbc122))
• {{embed ((69c1f832-a120-4831-94a4-d7ef9eecbc59))}}
• DONE Review Thuan’s work on DE-208

DONE Round 1

DONE Round 2

Done

• DONE Onboarding call 3 with Basata

• Shared use case of Date Drill

• Report issue for Auto mapping filter between viz and field

Good timing does matter

• Context: when I want to propose an idea to dbx team about setting up a Claude connector to interact with dbdiagram right inside claude desktop: Ampcode
• I can read it from each team’s sync to quickly decide whether to propose an idea
• For example, when I want to propose ((69c10475-619d-42df-b63e-1884b27ce548)), read the dbx team sync to find out whether they are having some similar key words in mind

dbt fails debugging process (collected from Thuan)

Tasks

Partly today

Done

• Coi lại đoạn gặp nhau sau 2 năm timeskip, ông già nấu ăn ở Water Seven (rất chill), arc người cá.
• Câu chuyện của băng hải tặc mặt trời rất cảm động.
• Tác giả xây dựng cái bất công của phân biệt chủng tộc. Trong truyện là con người và người cá.
• Mỗi lần người cá quyết định tin, thì con người lại làm họ thất vọng.
• Nhưng nhiều khi nguyên do cốt lõi chỉ đến từ “do chúng tôi không biết gì về các bạn”.

• Độ vẫn giữ như cũ, chỉ có đục thuỷ tinh thể trong mắt làm mình thấy mờ đi.
• Sẽ thay cái kính mới để chống ánh sáng mặt trời lúc ra đường.
• LEARNING nhớ nhỏ mắt thường xuyên, tránh để mắt khô.

Tasks

Partly today

Done

Tasks

Partly today

• ((69b3d030-98dd-450d-88d8-e00c1c322c59))
• ((69abeee3-b624-4e96-bd98-696afff2a17b))

Done

• ((69bcd2d3-e4f6-4661-95e4-80337127ac28))

• DONE Call with Add2Cart

• Remaining items:

  • How to refresh data (anh Dong handled).
  • How to set up other countries?
  • Prepare a PDF documentation to hand over.
  • Missing retailers in public.retailer. To do: follow up with Anurag.

Usually in data modeling-related tickets like this, to reproduce, we require:

• The how-to question.
• The dataset schema (which model, relationships, available dimensions/metrics).
• The expected AQL explore.

Usually in debugging a dbt test warning/failure:

• The error / signal.
• Which model? What are upstream models (to know the root cause can come from) and downstream models (to know affected models)?
• Query real data.
• Find out root cause. This may vary not only one root cause but many.
• Categorize root causes:
  • Data pipeline side
  • Source side
• Some common questions:
  • The test is testing what behavior? What does the test expect and not expect?
  • If we don’t feel the test make sense, question why is it there from first day?

In [[Oxygen Not Included]], there is a DLC that introduces a new kind of labor: Bionic Dupes.

[[1on1 Report - 2026-03-18]] notes

• Anh Triet’s concerns
  • AI (my primary focus on this)
    • How to integrate AI into development experience in data team?
    • What is required to make the team more productive?
  • Current data infra / pipeline being too complex
    • Anh Triet cant trust
    • Feeling it is being coupled into each other too much
  • PR Workflow (anh Triet feels no one is updating)

Recent thoughts about AI

• AI can make us worse but also can empower us

Tasks

Partly today

• TODO Think about applying AI into resolving #data-ops-bot issues (data pipelines)