Klai Focus
The confidential documents you couldn't put in ChatGPT. Now you can.
Upload contracts, reports, financial data. Ask in plain language. Every answer shows exactly where it came from.
Session documents
+ Add document
Session only. Documents are not stored after you close this tab.
From SPA_final_v3.pdf, Schedule 4:
The earn-out of €2.4M is payable if the company achieves revenue of at least €8M in FY2025. Payment is due within 30 days of the audited accounts being finalised.
SPA_final_v3.pdf · page 34, Schedule 4
Based on the management accounts, H1 revenue is €3.9M — running 2.5% below the pace needed. The accounts note a large contract expected to close in Q3.
management_accounts.pdf · page 3
How it works
Upload, ask, verify.
Upload your documents
PDF, Word, Excel, PowerPoint, plain text. Klai reads them privately, on EU infrastructure. Upload one document or twenty. Contracts, financial statements, technical manuals, due diligence files. Everything becomes instantly searchable.
Ask in plain language
No search queries, no Boolean operators, no reformatting your question to make a tool understand it. Ask the way you would ask a colleague who just read every document you uploaded. What are the payment terms? Which clause restricts assignment? Where does the liability cap kick in?
Answers with exact source references
Every answer shows the exact passage it came from. The page number, the paragraph, the quote. You are not asked to trust an AI summary. You see the source and you can verify it yourself. That is what makes it useful in professional settings where accuracy is not optional.
Capabilities
Citations, not guesses.
Always cited
Every answer references the exact passage it came from. Page number, paragraph, quote. You are never left wondering whether the answer is accurate or invented. If Klai found it, it shows you where. If it could not find it, it tells you that too.
Session-only, no retention
Documents are processed during your session and not stored afterwards. You close the tab and the content is gone. Nothing is indexed in the background, nothing is used for training, nothing is retained for a purpose you didn't ask for.
Entirely EU-processed
Your documents stay on EU servers the entire time. No US cloud provider handles them. No Cloud Act exposure. The companies running the infrastructure are based in Europe and operate under European law.
Any document type
PDF, Word, Excel, PowerPoint, plain text. Upload multiple documents and ask questions across all of them at once. Useful when you need to compare contracts, reconcile figures across reports, or find what multiple sources say about the same thing.
Who uses it
Research that stays between you and the document.
Search contracts without the risk
You have 30 supplier contracts and a client asking about liability exposure across all of them. You upload the stack, ask the question, and Focus finds every relevant clause with the exact source reference. No external party reads along. No document leaves your infrastructure.
Due diligence in a fraction of the time
A data room with 400 documents. A deadline in 48 hours. You upload what matters, ask the questions your client needs answered, and Focus pulls the relevant numbers and passages with citations. You spend your time on the analysis, not the searching.
Find what the manual actually says
The procedure manual is 300 pages. The engineer on-site needs to know the exact steps for a specific situation, right now. Upload it, ask in plain language, and get the relevant paragraph with the page reference. No searching, no scrolling, no guessing.
Your documents are private. Actually.
Most AI document tools keep your files to improve their service or train their models. Focus only processes your documents during your session. When you close the tab, they are gone. There is no retention, no indexing, no background processing you did not request.
Documents not stored after your session ends
No indexing, no training on your files
Processed on EU servers
Open source stack: every step auditable
Start using your private AI stack today.
Data never leaves the EU · Never trained on your data · Open source & self-hostable