Legal Decoder has launched Aperture, a natural-language interface for querying legal billing and spend data. The product is aimed at in-house legal teams and law firms that want to ask plain-English questions about invoices, staffing and matter efficiency without treating a large language model as the system of record.
The distinction matters in legal spend analysis, where a neat answer can still be useless if the path to that answer cannot be audited. Aperture sits on top of Legal Decoder’s analytics engine, which the company said has been validated in federal court proceedings and used across tens of billions of dollars in legal fees.
The product is built to ground conversational queries in structured billing analytics. Rather than summarizing invoices as blobs of text, Aperture uses Legal Decoder’s analytical framework to identify patterns in staffing, task execution, workflow efficiency and whether work matches the complexity of a matter.
Structured flags before chatbot answers
Legal billing data has long leaned on UTBMS codes, the task-based labels commonly used to categorize legal work. Aperture adds a more granular layer through 45 proprietary analytics flags developed from more than a decade of billing research and domain work, according to Legal Decoder.
That design puts Aperture in a specific corner of legal tech’s AI boom. A generic invoice chatbot can answer questions if the underlying text is available to it. Aperture’s claim is that the answer is constrained by a prebuilt analytical model, giving legal teams a way to review how a conclusion was reached and whether it rests on consistent billing signals.
David Solomon, Legal Decoder’s chief executive, said generative AI has made it easier to ask questions of data, while legal users still need transparency and confidence in the results. The company is positioning Aperture as a faster interface to billing intelligence, not as a replacement for the analytics layer beneath it.
Tokenization for sensitive legal data
The launch also puts data handling near the center of the product. Legal invoices can contain client names, matter references, strategy signals and other sensitive context that legal departments are unlikely to hand casually to an LLM.
Legal Decoder says Aperture applies tokenization before data reaches the model, replacing identifying markers with secure tokens. Joseph Tiano, the company’s co-founder and president, said the system resets tokens at the session level, so the same underlying data is not represented the same way across sessions.
That is the technical trust layer Legal Decoder is selling: structured analytics first, conversational access second, and token handling before model interaction. For buyers, the question will be whether those controls are enough to make AI-assisted billing analysis usable in workflows where auditability and confidentiality carry more weight than a slick demo.
Aperture is also intended for law firms that want to benchmark performance and support pricing strategies using matter data, as well as corporate legal departments looking for a more continuous view of billing behavior. Legal Decoder describes the Aperture legal billing intelligence platform as a way to turn invoice records into operational analysis, with defensible outputs tied back to its analytics engine.