Raj

Arjūn Rāj on X: "Made good progress with the @thegranturismo 7 telemetry project. Tracking car position along with throttle and brake input over the track in real-time and displaying with #SVG Green is throttle Red is brake and Grey is coasting If you look closely, you can see the upshifts and https://t.co/0vqe4tXrZZ" / X Made good progress with the @thegranturismo 7 telemetry project. Tracking car position along with throttle and brake input over the track in real-time and displaying with #SVG Green is throttle Red is brake and Grey is coasting If you look closely, you can see the upshifts and https://t.co/0vqe4tXrZZ

Granturismo 7 Telemetry app progress, built with #Tauri, #React and @shadcn.com/ui

How We Build at GoVisually: The May 2025 Edition A look at our evolving, AI-powered workflow for building GoVisually, from idea to release. This is how we work today, and how it keeps changing.

6. Staging and deployment
- By this time, the feature is tested multiple times
- AI does a final review on a release PR
- Release notes are prepared
- 🚀 And we ship

Read the full article on this here www.athousandnodes.com/posts/how-we...

5. Testing
- Manual testing happens on Preview apps and later on the Dev environment
- AI is really good at writing e2e tests as it's already aware of the scope of the feature and changes made
- Cypress does a full round of e2e

4. Code review
- @coderabbitai.bsky.social kicks in with its review on PR
- Developers also use #CursorAI to review PRs with @github.com MCP

3. Preview
- In development features are deployed to PR apps as preview apps
- Which can be tested in a sandbox on #Render

2. Starting the build
- Next, we would pull all this into #cursorAI
- Continue the research and planning in the project context
- And create markdown docs and push to a branch
- And start building

Here is how we are using AI in development at GoVisually.

1. Planning, research and Prototyping
- We start with simple requirements and start planning it using a project aware LLM
- Prototype it in @v0.dev
- Create spec and Issues on @github.com

Software engineering is costly. It requires Tokens.
#AI #VibeCoding

The future of label compliance is AI-powered, transforming tedious reviews into lightning-fast precision checks.

GoVisually's AI-assisted reviews are making this future a reality - catching costly mistakes in seconds.

Exploring a novel solution to render dynamic #react code in browser

I’m reviewing the documentation for
#stripe as I work on customer portals. Managing multi-product subscriptions is quite challenging. We have an API-based system that we want to transition to the customer portal.

Some experiments with #vercel ai sdk and #sandpack
- Product site with blog and auth pages
- Build with NextJS and Tailwind

Today recipe
- #GoVisually Playbook AI
- #Vercel AI SDK + Tool calling
- #Tldraw editor API
- And we mix 👨‍🍳
(Analyzing proofs and annotating remarks on them)