graham walker

[graham walker @gwalker]

@everestchris6 video

[graham walker @gwalker]

@tomcrawshaw01 Agent

[graham walker @gwalker]

@reachvaldo AF

[Connor Davis @connordavis_ai]

Agentic RAG is the upgrade everyone pretended we didn’t need… until it steamrolled vanilla RAG.

RAG owned 2023. Then 2024 hit and builders realized something obvious: retrieval alone isn’t enough. The future belongs to systems that can route, plan, validate, and correct themselves while pulling info from multiple sources. That’s where Agentic RAG comes in.

Think of it this way.

→ Vanilla RAG is a library card.
→ Agentic RAG is a smartphone with unlimited tools.

Here’s the real idea behind Agentic RAG:

You don’t just bolt retrieval onto an LLM. You give the LLM an agent that can think through the problem, choose the right tool, pull data from multiple sources, evaluate what it found, and try again if the context is weak.

The agent becomes the brain of the pipeline.

A basic RAG pipeline hits two hard walls:

• It only pulls from one source
• It retrieves once and hopes for the best

Agentic RAG blows past both.

An agent can choose between a vector index, a web search, a calculator, an API, or anything else you plug in. It can rewrite the query, re-run retrieval, compare sources, and validate the results before handing anything to the model.

That’s how you jump from “here’s some context” to “here’s the answer you actually needed.”

The architecture can stay simple one agent acting as a router or grow into a multi-agent setup with specialists: one for internal data, one for emails, one for the web, one for external APIs. The master agent orchestrates the whole thing.

This is why enterprises are moving fast.

Developers get copilots that fix their own mistakes. Business teams get agents that can search across tools automatically.

Retrieval becomes a reasoning loop, not a single step.

Agentic RAG delivers better context, better accuracy, and far more autonomy. The trade-offs are real: more latency, more points of failure, and more dependence on the LLM’s reasoning ability. But the gains are too big for serious teams to ignore.

Vanilla RAG was phase one.

Agentic RAG is the first real step toward AI systems that can work through problems instead of just answering them.

[Karl Mehta @karlmehta]

He predicted:

• The Deep Learning revolution (2008)
• The online education boom (2011)
• China's massive AI dominance (2014)

Now Andrew Ng revealed 5 opportunities that will create more millionaires than anything before.

Here's what you should know (& how to prepare): 🧵

[graham walker @gwalker]

@FynCas MASS

[Cas.Fyn @FynCas]

We’re now generating 100+ videos per minute.

This is mass AI marketing — and it changes everything.

• UGC creative costs? Slashed
• Video production? Instant
• Reach? 10x scale overnight

This isn’t the future.
It’s already happening.

RT + comment “MASS” and I’ll show you how we’re doing it.
(must be following)

[graham walker @gwalker]

@SubhaghV Call

[Subhaghvs @SubhaghV]

Just built a Voice AI Chatbot that lets customers talk to your business and get real answers from your docs.

Sold the setup for $3,000 + $750/mo and it’s already answering calls for multiple businesses

Here’s what it does:

– Lets users ask questions via voice (like: “Do you have vegan options?” or “Are you open this Sunday?”)
– Understands natural speech via ElevenLabs
– Pulls answers directly from your docs using Qdrant + OpenAI (RAG-based)
– Replies back in real-time — with a human-sounding voice
– No retraining. Just update your FAQ/Docs in Google Drive and it works
– Optional: Embed it on your site to replace live chat or IVR

Perfect for restaurants, salons, clinics, retail stores any business getting the same 10 questions every day.

Follow + RT + Comment "CALL"

I’ll DM you the full plug-and-play workflow with setup instructions.

[graham walker @gwalker]

@SubhaghV Chat

[Subhaghvs @SubhaghV]

Just built a WhatsApp AI agent that replaced an entire customer support team.

Sold the setup for $2,000 + $500/mo and clients are begging for more.

Here’s what it does:

– Reads your company’s Google Doc (FAQs, policies, service info)
– Lets customers ask anything on WhatsApp
– Replies instantly using ChatGPT or Gemini
– Handles hours, pricing, bookings, even dynamic logic like closure dates
– No retraining, no dev work just update the Doc and it works

Perfect for gyms, clinics, restaurants, service businesses, or anyone tired of answering the same 20 questions daily

Built it in n8n + WhatsApp Cloud API: no code, no friction.

Follow + RT+ Comment “CHAT”

I’ll DM you the entire plug and play workflow

[graham walker @gwalker]

@kirso_ @iannuttall Interesting observation. Do you use a PRD template for Gemini or are you referring to the CLI?

[graham walker @gwalker]

Game changer in every way. Amazing to me that less than 5% of businesses know what @lovable is.

[Anton Osika @antonosika]

Lovable just crossed $100M ARR in 8 months.

Faster than OpenAI, Cursor, Wiz, and every other software company in history.

Today we're launching a game-changing update, it reduces the error rates by 91%.

Introducing Lovable Agent:

[graham walker @gwalker]

@TheVeller Plan

[graham walker @gwalker]

@builtbysegdae @NotionHQ thanks for sending... looks great

[graham walker @gwalker]

@builtbysegdae AI

[Segdae - The AI FDE @scalewithshay]

AI will make you superhuman.

But most people don’t know where to start.

That's why I built the Ultimate AI Toolkit:

• 600+ leading AI tools
• Books
• Prompt guides
• Newsletters & more.

Free for 48hrs.

To get it, just:
1. RT
2. Follow me (for auto-DM)
3. Reply "AI"

[graham walker @gwalker]

@yaiyi66 @ssocialtrading @drchethansathya https://t.co/qO9OQJ7oTI
#CCCA

[graham walker @gwalker]

Evidence that demands a response.
#CCCA
https://t.co/qO9OQJ7oTI

[graham walker @gwalker]

@ACLU @maxwellsmith vaccinated people are 13-27 times more likely to contract covid and 8 times more likely to be hospitalized when compared with people who have natural immunity. It almost appears as if the mRNA vaccines make people MORE susceptible to the virus rather than less susceptible