Legal tech company A2D2 launched Sherlock, an evidence analysis chatbot that turns financial transaction records into court-ready evidence, at the 2026 Korean Legal Industry Fair in Seoul from the 24th to the 26th. The tool targets a routine but costly bottleneck in litigation: manually copying hundreds of bank transaction lines, deposit slips, and card statements into Excel.
Sherlock's core capability is extracting and standardizing financial documents from entire case files. When an attorney uploads a case record, the AI automatically identifies financial documents mixed in with other materials. Scanned PDFs with hard-to-recognize text are read via optical character recognition (OCR) and then converted into a uniform table showing transaction dates, amounts, withdrawals, and deposits. Different banks and companies use their own forms; Sherlock unifies these formats so that lawyers can download hundreds of pages into a single Excel file in one second.
The tool does not stop at transcription. It tracks the flow of money across multiple people's account statements and labels each transaction with an evidence priority: core, important, or reference. This grading lets lawyers immediately see which deposits or withdrawals are decisive for the case. Every converted table remains linked to the original document, preserving the chain of accuracy that courts require.
Conversational evidence review
The entire analysis unfolds through a chat interface. An attorney types a prompt like "Find the remittance details the day before the incident" or "Financially analyze the transaction details," and Sherlock surfaces the relevant data from the case record and cites the underlying evidence. No keyword searches or complex menu navigation are needed.
"The analysis of financial transaction details in civil lawsuits is one of the most painful tasks for lawyers," A2D2 CEO Kim Yoon-woo said. "Irex automates this complex process, providing an environment where lawyers can clearly identify the flow of money and focus only on the nature of the case and winning strategies."
Integration with the broader Irex platform
Sherlock connects to the existing evidence analysis functions of Irex, A2D2's larger legal AI system. Irex can analyze up to 100,000 pages of evidence at once, summarize key issues, and automatically organize witness statements and event timelines. When a lawyer highlights a critical piece of a document, Irex generates a draft preparation document and pulls up relevant laws and precedents. The combination links the earliest stage-untangling scattered financial records-to the final stage of drafting pleadings.
Since Irex launched in July of last year, more than 235 law firms have adopted it, with 535 registered cases and over 942,000 pages of evidence uploaded. AI for Legal tools that automate document review are becoming standard equipment for litigation practices, and Sherlock extends that capability deeper into financial forensics.
Why this matters for legal professionals
Financial document analysis in disputes-fraud, embezzlement, breach of trust-often means days of manual data entry. By converting that work into a conversational, automated process with built-in evidence grading, Sherlock lets lawyers and paralegals redirect their time toward case strategy and argument construction. Paralegals who manage large-scale document reviews can explore similar time-saving techniques through an AI Learning Path for Paralegals. For law firms handling client funds or complex commercial litigation, tools that reliably trace money flows and link them directly to source documents reduce both human error and preparation time.
Your membership also unlocks:
