Selling AI chatbot builds on a Korean freelance marketplace, alongside a full-time job
A one-person listing that sells internal-document chatbots as fixed packages on Kmong, Korea's largest freelance marketplace. No revenue figure is published anywhere, so we do not print one. What we do print is the platform-displayed transaction count and the published price list — the price band is the useful part here.
- Founder
- Nunnopi Coding (Kmong seller name)
- Country
- KR
- Team
- 1
- Startup cost
- —
- To first revenue
- —
- Automation
- mid
- Reported revenue
- Revenue not disclosed
What is sold
Three fixed packages, separated by scope, turnaround and revision count
STANDARD ₩150,000 (3 days, 2 revisions) / DELUXE ₩350,000 (5 days, 3) / PREMIUM ₩700,000 (7 days, 5). Checked 2026-08-15.
Who buys
Companies automating internal document search and customer replies
AI's part and the person's part
Already had
MSc in computer science. Machine-learning engineer, then data scientist, then analytics consultant, and VP of Engineering since February 2025 according to the profile. This is a side listing run alongside a full-time role.
If this is large, the AI is not the reason
What the AI did
Builds chatbots that search and summarise a company's own documents and interpret internal policy. The platform lists the tools used as ChatGPT, Midjourney, DALL·E and Stable Diffusion. What is sold is not AI itself but the configuration against one company's documents.
the part a tool replaced
What the person did
Requirement interviews, workflow design, document indexing and embedding settings, delivery and a usage guide. The listing says structure matters more than document volume — setting up that structure is the human part.
stories that omit this do not replicate
Tools: chatgpt · midjourney · dall-e · stable-diffusion
How customers were found
Kmong — marketplace search placement
Platform-displayed “replies within 3 hours” (a ranking signal)
Paying customers
STANDARD ₩150,000 (3 days, 2 revisions) / DELUXE ₩350,000 (5 days, 3) / PREMIUM ₩700,000 (7 days, 5). Checked 2026-08-15.
Flow: Kmong — marketplace search placement, then Platform-displayed “replies within 3 hours” (a ranking signal), then Paying customers
How we verified this
Public dashboard or platform figure. May be self-reported
Checked the Kmong listing directly on 2026-08-15. Platform-displayed values — 18 completed transactions, 96% satisfaction, individual (not corporate) account, team size 1. The price list is published on the page and copied as-is. ⚠️ Those 18 transactions are the seller's total across all listings; this chatbot listing itself is new and has zero reviews. No revenue figure is published anywhere, so we print none — multiplying transactions by price would be our estimate, not a fact.
- Kmong listing — transactions, satisfaction, team size and prices (marketplace · checked 2026-08-15)
What made it work
What is sold is the setup, not the AI. Buyers are not stuck because they cannot use ChatGPT; they buy so they do not have to fit it to their own documents. That is also why the price list is split by scope, turnaround and revision count rather than by outcome — the product is a bounded amount of labour.
Risks
⚠️ The expertise bar is high. A master's degree and a work history are doing the trust work here; the same prices are hard to charge without them. ⚠️ We cannot confirm this listing sells at all — it has zero reviews, and the 18 transactions belong to other listings. How long a new listing takes to get its first order is not something this case answers.
Does this work in Korea?
This is a domestic case, so there is little friction. Kmong accepts individual accounts and handles payment and settlement. Platform fees and income reporting are the seller's own; whether they can issue a tax invoice matters to corporate buyers — this seller is marked as unable to.
Solo-feasible: yes · Automation: mid
7-day action plan
Step 1. Search a freelance marketplace for “AI chatbot build” and copy the price lists of the top ten
Step 2. Tabulate what those ten have in common: document integration, revision count, turnaround
Step 3. Pick the smallest scope you could deliver within three days
30-day action plan
Step 1. List that scope once, priced near the low end of those ten
Step 2. Log how long you take to answer each enquiry
platforms rank on it
Step 3. After the first order, measure the hours you actually spent and divide. If the hourly rate is poor, cut the scope
Run the numbers
We do not print a single projected figure here. Put your own numbers in instead.