A private AI platform for a Bangkok video production company. Every past invoice, quote, treatment and supplier cost now sits in one knowledge base — and a proposal generator turns a brief and a budget into a priced budget in the company’s own design.
| Item | Median | Range | Jobs |
|---|---|---|---|
| DOP, per day | ฿28,000 | 18–45k | 42 |
| Camera package | ฿22,500 | 15–38k | 37 |
| Gaffer and grip | ฿16,000 | 12–24k | 51 |
| Studio, per day | ฿35,000 | 20–60k | 29 |
| Colour grading | ฿18,000 | 12–30k | 24 |
The challenge
A proposal for a large production took days, sometimes a week. The prices were all there — spread across ten years of invoices, quotes, supplier costings and project folders — but nobody could find them quickly. The team opened old files, compared spreadsheets and called suppliers to confirm what a day of crew or a studio really cost.
Our approach
We collected everything from the past ten years: invoices, quotes, project data, storyboards, treatments, supplier costs and creative work. Over two weeks we sorted it with the team and loaded it into one database. Then we extracted the information into short, structured snippets, so the AI reads a clean knowledge base instead of thousands of old documents every time someone asks a question — fast answers, low AI costs.
What we built
Ten years of invoices, quotes, project data, storyboards, treatments, supplier costs and creatives, sorted and loaded into one database.
The information extracted into small, structured snippets, so the AI finds answers without re-reading old documents every time.
Prices estimated from past jobs and proposals, with the median, average and range for every line item.
Takes the project requirements and a budget, then prepares a full budget from the company’s own price data, in its own proposal design.
Bangkok Productions
Why it pays off
Most companies already own the data that would make them faster: years of quotes, invoices and project files. Organised into a knowledge base, that history becomes an asset that works every day. Proposals that took days now start from a complete, priced draft, and the team can analyse ten years of jobs — costs per category, supplier trends, budgets per project type — in ways that were never possible before. Once a workflow like this is in place, every process built on it speeds up, and every new job makes it more accurate.
Your old invoices and quotes hold the answers to what things cost and who did the work. Here is a seven-step method to make them searchable and useful to AI.
Using AI on your own quotes, supplier prices and customer records raises fair questions about privacy. Here is where the data goes and how to keep control.
Not every process deserves AI. Here is how to spot the ones that do, score them in an afternoon and start with a small, safe first project.
LINE apps, web platforms and AI features — tell us what you need and we will reply in person.
Start a project