How much does AI consulting cost?

BlokAI's engagements range from $5,000 to $75,000 depending on scope. Simple automations and LLM integrations start around $5K–$15K. Custom AI agent systems and full data infrastructure projects run $30K–$75K. We scope every project before committing to a price.

What does BlokAI build?

BlokAI builds AI agents, LLM applications, RAG systems, automation pipelines, and data infrastructure for startups and SMBs. We take projects from architecture to production — not demos, not prototypes.

Is BlokAI based in Japan?

Yes. BlokAI is a solo studio run by Kent Tran, based in Tokyo, Japan. We work with clients globally in English.

What is RAG and why does my business need it?

RAG (Retrieval-Augmented Generation) lets AI systems search your own documents, databases, or knowledge bases before answering — instead of relying only on training data. It powers AI customer support, internal search tools, and sales enablement systems that are accurate and up-to-date.

ChatGPT Is Not an AI Strategy: What Businesses Actually Need in 2023

"ChatGPT has reached 100 million users just two months after launch — the fastest-growing consumer application in history." — Reuters, February 1, 2023

It's been three months since ChatGPT dropped and every business owner I talk to has the same look on their face. Half excitement, half panic. They've seen the demos. They've read the headlines. And now they're in my inbox asking some version of: "How do we get this into our business?"

Fair question. Wrong framing.

ChatGPT is a product. It's a wrapper around a language model built by OpenAI, optimised for consumer use, and free to sign up for. It's genuinely impressive. But it is not an AI strategy any more than "using Google" is a search strategy.

Here's what I'm seeing on the ground, and what we're advising clients at BlokAI right now.

The Tool Confusion Problem

When a new piece of technology gets this much attention this fast, the first instinct for most businesses is to ask: "How do we use this tool?" But that's the wrong starting point. The right question is: "What problem are we solving, and is AI the right approach?"

Most businesses right now are doing one of three things:

Nothing — waiting to see how it shakes out (understandable but increasingly risky)
Experimenting randomly — employees using ChatGPT for ad-hoc tasks, no coordination, no data strategy
Chasing the headline — leadership wants an "AI project" without knowing what success looks like

None of these is a strategy.

What a Real AI Strategy Looks Like

We've been doing AI implementation work long enough to know that the projects that succeed share a few common traits:

They start with a specific workflow, not a general capability. Don't ask "how can we use AI?" Ask "which of our current workflows involves repetitive text generation, classification, or retrieval?" That's where LLMs actually help. They have a clear data story. Language models are powerful, but they're not magic. If your business runs on proprietary knowledge — internal docs, customer histories, product specs — that data needs to be surfaced, structured, and connected to the model. This is where most DIY attempts fall apart. They define what success looks like before they start. Is it time saved per employee? Reduction in support tickets? Faster report generation? If you can't measure it, you can't improve it. They treat AI as infrastructure, not a feature. The companies that will get lasting ROI from AI are the ones building it into their workflows at the infrastructure level — not bolting on a chatbot as an afterthought.

What We're Actually Recommending Right Now

At BlokAI, our current advice to most SMB clients is: don't integrate ChatGPT the product — think about integrating the capability.

That means looking at the OpenAI API (or alternatives), and thinking about where language understanding and generation can replace manual steps in your existing processes. Customer support triage. Document summarisation. Internal Q&A. First drafts of proposals. Meeting notes into action items.

These aren't glamorous use cases. But they're high-frequency, they're measurable, and they add up quickly.

The other thing we're recommending: protect your data from day one. Sending sensitive business data to consumer AI products without understanding the terms of service is a liability you don't want. Enterprise agreements, private API access, and local deployment options (more on those in future posts) are all worth understanding before you scale up.

The Japan Context

Working from Tokyo, I see an additional layer of complexity for businesses operating here. Enterprise software adoption tends to be slower and more deliberate — which is actually an advantage right now. There's more room to get this right. Japanese enterprises are asking hard questions about data sovereignty, vendor lock-in, and long-term reliability. Those are the right questions.

For SMBs in Japan and globally, the opportunity is real. But it requires being intentional.

The Actual First Steps

If you're reading this and wondering where to start, here's a practical sequence:

Identify 3 workflows in your business that involve processing, generating, or retrieving text
Estimate the time cost of those workflows per week across your team
Check your data — what information would an AI assistant need to be useful in those workflows?
Talk to someone who's built this before — not to get sold something, but to understand what's realistic

The businesses that will be ahead in 12 months aren't the ones who moved fastest in February 2023. They're the ones who asked the right questions now and built on solid foundations.

Want to implement AI in your business? Book a 30-minute architecture call at blok-ai.com.