Abid Gogi

🤖 AI & Machine Learning Enthusiast
☁️ Cloud Computing
💻 Aspiring Tech Professional
🚀 Learning • Building • Growing

[Abid Gogi @AbidGogi]

Anthropic just ditched Git for Scion—their custom VCS built for AI agents

Git was built for offline humans holding full clones

Scion is built for always-online parallel agents doing 30k+ server-side merges daily

When agents write most of the code, version control has to evolve

[Abid Gogi @AbidGogi]

Anthropic just ditched Git for Scion—their custom VCS built for AI agents

Git was built for offline humans holding full clones

Scion is built for always-online parallel agents doing 30k+ server-side merges daily

When agents write most of the code, version control has to evolve

[Abid Gogi @AbidGogi]

The biggest mistake "vibe coders" make: ignoring authorization.

Your AI can scaffold authentication in seconds and still leak another user's private data. 😳

🔐 Authentication = Who you are

🛡️ Authorization = What you can access

Always test with 2 accounts before shipping. 🛠️

[Abid Gogi @AbidGogi]

The biggest mistake "vibe coders" make: ignoring authorization.

Your AI can scaffold authentication in seconds and still leak another user's private data. 😳

🔐 Authentication = Who you are

🛡️ Authorization = What you can access

Always test with 2 accounts before shipping. 🛠️

[Abid Gogi @AbidGogi]

@codvynn Hey! Appreciate the shoutout. Right now experimenting with health assistant workflows and conversational RAG—focusing on clean chunking so the model doesn't hallucinate.What are you building lately?

[Abid Gogi @AbidGogi]

Hit 21 followers! 🎉

Looking to connect with more builders:

🐍 Python & Backend
🤖 AI & RAG
🚀 Freelancers & Indie Hackers
🎯 Freshers on the job grind

Say hi, drop what you're building, and let’s connect! 🤝

[Rahul Mirji @MirjiRahul]

Day 01/60: System Design 🔥

Why do apps crash? One server. 🖥️

➕ Follow so you don't miss Day 2: what actually happens when you type https://t.co/FRLLDEpNwU and press enter?

[Rahul Mirji @MirjiRahul]

Hey @X Connect me with ⭐️

- Vibe coders
- Developers
- AI Engineers
- Coders

Let's connect.

[Abid Gogi @AbidGogi]

AI has commoditized writing syntax, but it cannot decide how your system handles failure.

As code generation becomes free, the entire value of an engineer shifts to: • System boundaries • Data consistency • Cost & latency trade-offs
System design is the moat. 🛠️

[Abid Gogi @AbidGogi]

AI has commoditized writing syntax, but it cannot decide how your system handles failure.

As code generation becomes free, the entire value of an engineer shifts to: • System boundaries • Data consistency • Cost & latency trade-offs
System design is the moat. 🛠️

[Abid Gogi @AbidGogi]

The biggest shift in dev tooling isn't autocompleting functions—it's agentic CI/CD.

Offload linting, basic test coverage, and regression checks to automated AI passes.

Humans should review architecture and risk tiers, not missing commas. 🛠️

[Abid Gogi @AbidGogi]

The biggest shift in dev tooling isn't autocompleting functions—it's agentic CI/CD.

Offload linting, basic test coverage, and regression checks to automated AI passes.

Humans should review architecture and risk tiers, not missing commas. 🛠️

[Abid Gogi @AbidGogi]

Hit 50+ followers! 🎉

Looking to connect with more builders:

🐍 Python & Backend
🤖 AI & RAG
💻 Next.js & Full-Stack
🤝 Freshers shipping projects

Drop what you’re building below—let’s connect! 🤝

[Abid Gogi @AbidGogi]

Hit 50+ followers! 🎉

Looking to connect with more builders:

🐍 Python & Backend
🤖 AI & RAG
💻 Next.js & Full-Stack
🤝 Freshers shipping projects

Drop what you’re building below—let’s connect! 🤝

[Abid Gogi @AbidGogi]

@MirjiRahul 😂

[Abid Gogi @AbidGogi]

@bigdata_engnr That’s a powerful combo. Are you working mainly with Superset / Metabase on the BI side, or plugging LLMs directly into custom text-to-SQL layers?

[Abid Gogi @AbidGogi]

@halalstocktok Haha 100%. "I take it back, you were innocent, it was just my terrible query plan." 💀

[Abid Gogi @AbidGogi]

The 3 stages of debugging an API endpoint:

1. "Why is this taking 3.8s? The model must be throttling."
2. "Wait, the external call took 200ms."
3. Discovering an unindexed SQL query running inside an async loop. 💀

Tech stack: 1, Ego: 0.

[Abid Gogi @AbidGogi]

@incheif01 Hlo

[Abid Gogi @AbidGogi]

@bigdata_engnr Followed back! 🤝 What data stack or pipelines are you currently working with?

[Abid Gogi @AbidGogi]

@TejasRehangdale Appreciate the connect, Tejas! Followed back. What tech stack or project are you currently focusing on? 🚀

[Abid Gogi @AbidGogi]

@jevlorde Hey Jev, followed back! AI agents on Android is an awesome intersection. Are you running on-device models with MediaPipe / LiteRT or hitting cloud endpoints? Happy to share RAG notes anytime! 🤝