Hannune

I automate the document work your team still does by hand. Translation, extraction, review. The formatting survives, and the numbers get checked.

8 years

Production AI experience across enterprise clients

5 hrs → 40 min

Document search time for a tax firm

2 patents

AI optimization patents registered in South Korea

Recent Work

Case 01

Boditech Med · Translation pipeline

IFU localization · pilot in progress · 2026

Formatting preserved Terminology control Translation memory

A docx-to-docx translation pipeline I built and run. Tables, numbering, and styles come back where they were instead of collapsing into plain text. Approved terminology is enforced from a glossary, and the terms that must never be translated are locked before the model sees them. Prior approved sentences are reused through a translation memory. Adding a language does not mean rebuilding the pipeline.

Case 02

Corporate training

In-house engineering training · 2026

8 sessions Hands-on Corporate training

Eight sessions of hands-on AI engineering training for an in-house committee at a medical device manufacturer, run on their own working problems rather than toy examples. Completed.

Case 03

Gahyun Tax · Document search

SMB · Korean tax domain

5 hrs → 40 min Korean NLP RAG

A tax practitioner was spending up to five hours hunting through tax code, precedents, and old case files for one answer. That is now under 40 minutes, and the answers are more often right. A retrieval pipeline over the tax code and case files is what made it work; the client only cares that the search stopped eating the day.

Case 04

LG Energy Solution · Multi-agent pipeline

Enterprise · 2026 Q1

LangGraph Crawl4AI EXAONE Multi-agent

A LangGraph multi-agent pipeline for talent data acquisition across global sources, with Crawl4AI for structured extraction, EXAONE for domain reasoning, and an evaluation framework to keep the output honest.

Services and Products

01 · Service

Available now

Document Automation

Translating, extracting, reviewing, and converting the documents a company runs on. Two things I care about more than anyone else seems to: the file comes back looking like the file you sent, and the figures I pull out are checked against each other before you see them. Fixed scope, fixed price, fixed timeline.

See packages ↓

02 · Product

Waitlist · Q3 2026

Entity Resolution API

Entity Resolution as a service. Deduplication and entity linking across datasets — built on Splink with an LLM disambiguation layer. No self-hosting required. Subscribe from $19/mo.

Visit app.hannune.ai →

03 · Product

New · Beta 2026

Ownership API

Northeast Asia 5%-rule and acquisition disclosures from Korea, Japan, China, Hong Kong, Taiwan, and US ADRs unified into one cross-lingual ownership graph. Built on Hannune ER API. From $49/mo, around one tenth of comparable enterprise alternatives.

See Ownership API →

04 · Platform

Live now

2asy.ai

Northeast Asia (Korea, Japan, China, Hong Kong) risk and compliance intelligence. It reads economic and regulatory disclosures daily, extracts causal relationships, and publishes structured analyses. Same document machinery, pointed at public filings instead of a client's files.

Visit 2asy.ai →

01 · Service Available now

Document Automation

Most tools translate a document well and hand it back with the table broken and the columns collapsed. Most extraction tools pull numbers out and never ask whether the line items add up to the total on the summary page. That gap is the work: your files come back in their original shape, and the figures are reconciled before anyone reads them. Fixed scope, fixed price, fixed timeline, and you deal with the engineer who builds it. Your documents stay wherever you decide, in your cloud account or on your own servers.

Package 01 · Validate

Starter

One document type, run end to end on your real files before you commit to anything bigger.

$8,999 fixed

3 weeks · Fixed scope

  • ScopeOne document type: your files in, finished files out, formatting intact
  • FitsTranslation · extraction · summarizing · format conversion
  • OutputWorking pipeline + an accuracy report on your own documents
  • Incl.A straight answer on what is worth automating next

Package 02 · Build

Standard

The whole document workflow in production, running on its own.

$38,999 fixed

6 weeks · Fixed scope

  • ScopeSeveral document types, batch volumes, deployed and running
  • ChecksTerminology rules, do-not-translate lists, totals reconciled across sections
  • OutputProduction system + 1 month hypercare
  • Incl.Full codebase transfer + docs + training
  • SLA6-week delivery guarantee

Package 03 · Scale

Enterprise

Several teams, several document types, inside your own infrastructure.

Contact for pricing

Typical range $100K–$250K+

12+ weeks · Custom scope

  • ScopeDocument workflows across teams, plus search and review on top of them
  • RunsYour cloud account, your own servers, or fully air-gapped — your call
  • OutputFull system + 3 months on-call support
  • Incl.Retrieval and agent layers where they earn their keep · custom model tuning
  • SLADedicated support

Add-on · Ongoing

Ops Retainer

Keep the pipeline accurate as your documents change.

Available after any completed package. Includes: a monthly accuracy pass on recent output, new templates and terminology added as they appear, model updates, priority Slack support, and up to 8 hours of additional development per month.

From

$3,499 /mo

Min. 3 months

Custom engagements

Retainer-only arrangements, training programs, or scope beyond the three packages. We reply within 2 business days.

Request a call

Live Demos

Three working systems you can use right now. Upload one of your own PDFs to the last two and see what comes back.

Document Q&A

Upload a PDF and ask it questions, back and forth, without reading the whole thing yourself.

PDF Upload Retrieval Multi-turn

Document Extraction

Upload a PDF and get the fields back as structured JSON instead of retyping them.

PDF Upload Structured Output JSON

Ask Hannune

A multi-turn assistant that answers questions about Hannune from a structured profile.

Multi-turn Profile-grounded Conversational

Long Documents, Read Properly

Three public documents nobody wants to read end to end: a 418-page competition ruling, a 187-page trade review, and an SEC annual filing. Each one was put through the same pipeline and came out as something you can actually follow. Knowledge-graph retrieval sits underneath, which matters to me and not to you.

What I Build

Four kinds of document work, and the technology I reach for underneath them.

The document work

Translation, format intact Extraction to structured data Summarize & review Document-to-document conversion

The two hard parts

Layout & tables preserved Terminology & do-not-translate rules Totals reconciled

Underneath, when it earns it

Graph RAG, Neo4j LangGraph multi-agent Cross-language entity resolution Docker & REST, your environment

2asy.ai: Northeast Asia Risk and Compliance Intelligence

The same document machinery, pointed at public filings instead of a client's files, and left to run on its own. Every day it reads economic and regulatory disclosures across Korea, Japan, China, and Hong Kong, pulls out the causal relationships, and publishes structured analyses.

It happens to be tracking oil, energy, and geopolitics. The pipeline does not care about the subject; point it at a different pile of documents and it works the same way.

2asy.ai preview showing knowledge graph and analysis
  • ~2,400 articles ingested daily
  • ~100 articles selected for knowledge graph
  • Causal graph visualization on every edition
  • Runs on Neo4j + LangGraph + Ghost CMS

Identity

One person, accountable from the first conversation to the system running in production. No account manager in between.

  • Build and deliver the system end to end
  • Own the architecture, the code, and the deployment directly
  • Contract work for companies and professional firms
  • Solo by design — one person to ask, one person to blame

Operating Model

One technical owner from system design to deployment, with no handoff gaps between planning, implementation, and operation.

Fit

This works best when there is a real pile of documents and a real person losing hours to it.

When this is a good fit

Your documents have structure worth keeping — tables, columns, numbered sections, house terminology — and the off-the-shelf tools keep wrecking it. Or you are pulling figures out by hand because nobody trusts an extraction they cannot check. Or the same conversion happens every week and someone is doing it by hand.

When this may not be a fit

A handful of plain documents a month that ChatGPT already handles fine. Research projects with no workflow behind them. And anything where nobody has actually looked at what the current process costs.

Get In Touch

Send me a document your current tools handle badly and tell me what it should have looked like. That is usually enough for me to say whether this is worth your money.