In-house AI
An in-house LLM with document search and Q&A, running on your own server. Data never leaves the company, and usage adds no per-call cost.
Who this fits
- Companies where a person types the same answer to the same question over and over
- Organizations sitting on rules, manuals, and past quotes they cannot find when needed
- Companies that want AI but are uneasy about uploading internal documents to an outside service
How far we take it
Included
- Deciding which questions it should answer
- Collecting internal documents and building a search index
- An answer view that shows the source documents alongside the answer
- Integration with tools you already use (messenger, internal systems)
- A way to check answer quality and a process to improve it
- Operations documentation and handover
Not included
- Having a person read, organize, and write up your scattered documents
- Model usage fees and infrastructure costs such as GPU servers
- Legal or accounting review of what the answers say
- Company-wide training or an on-site presence after rollout
Build cost by scope
AI chatbot & automation2~6 weeks
| Scope | Build cost |
|---|---|
| Start smallCore features only | ₩2M~5.2M |
| StandardTypical scope | ₩4M~8M |
| LargeMany screens, complex | ₩6.4M~10M |
On-premise LLM1~3 months
| Scope | Build cost |
|---|---|
| ValidationSingle model & inference API | ₩10M~18M |
| DepartmentDepartment auth & usage controls | ₩15M~25M |
| EnterpriseMulti-node & high availability | ₩21M~30M |
These are example ranges. Once requirements are confirmed you get a fixed quote. Maintenance is billed separately as a monthly plan.
These figures are current as of September 2026. The latest ranges are on the pricing page.
How it runs
- 01
Discovery
We define the core user problem and align on scope and priorities.
- 02
Design
We design the structure and flows, validating fast with prototypes.
- 03
Build
We build maintainable systems with tests and code review.
- 04
Launch
We ship reliably with release automation and store submission.
- 05
Operate
We operate long-term with observability, logs, and metrics.
Frequently asked questions
It depends which route you pick. With an external model, the question and parts of the matched documents are sent to that service. An on-premise LLM keeps the model on your own servers, so documents never leave and there is no per-call cost that grows with use, at the price of upfront cost and hardware. We decide together based on how sensitive the documents are and how much you will use it.
We build assuming it will be wrong sometimes. It does not just produce an answer; it shows which document and which passage it relied on, so a person can check on the spot. That is what makes it usable for real work. When it cannot find grounds, we tune it to say it does not know rather than invent something. This is the deciding factor for adoption, so we design for it from the start.
You can. We just do not load everything. We pick around ten questions people actually ask, then feed in only the documents that really contain those answers. That reveals which documents are useful and which are out of date, and the tidying tends to happen along the way. Waiting until everything is organized usually means never starting.
The first fork is whether documents may leave the company. If they may, the chatbot route starts faster and cheaper. If they may not, it is on-premise. The second is volume: external models charge per call, so past a certain daily usage on-premise becomes cheaper. In the first conversation we look at the nature of your documents and expected volume and point out where that crossover sits.
Related public docs
Proposals, scope guides, and build examples we publish openly.
A vague scope is fine
Tell us where things stand and we will point to the scope you need and a rough size.