Getting the Most Out of LexUnits: Tips, Shortcuts, and Hidden Features
LexUnits power user tips are workflow optimisations that help Australian and New Zealand lawyers generate higher-quality billing entries with fewer credits and less manual editing. Most users start with the basics — upload a recording, review entries, export — but the platform has depth that is not immediately obvious. This guide covers the techniques that experienced users have found make the biggest difference to output quality and time savings.
According to internal usage data, lawyers who use Session Context pre-fills and AI Polish together spend an average of 40% less time editing entries compared to those who generate without context. The difference is significant: approximately 3 minutes of review time versus 7 minutes for a typical 10-entry batch. Over the course of a month, that adds up to hours of recovered time. These tips are drawn from real user workflows and are arranged from highest-impact to more niche optimisations.
Tip 1: Always Fill in Session Context
The single highest-impact habit is filling in the Session Context panel before clicking Generate. The panel has four fields: Matter, Date, Timekeeper, and Client. When these are populated, the AI uses them as anchor data — every generated entry includes the correct matter number, the correct date, and references the client by name.
Timekeeper auto-rate matching: When you enter a timekeeper name, LexUnits automatically applies the correct billing rate. If your rates list includes "William — $700/hr" and you type "William" in the Timekeeper field, every generated entry uses the $700 rate. This works with exact name matching first, then fuzzy matching for common variations (e.g., "Will" matches "William"). This feature alone eliminates the most tedious manual edit — fixing the rate on every entry.
Session Context persists within a session, so if you are processing multiple recordings for the same matter, fill it in once and it applies to all subsequent generations until you change it.
Tip 2: Use Polish Aggressively — It's Free
AI Polish (the ✨ sparkle icon) is the most underused feature in LexUnits. Polish refines billing descriptions — improving language, expanding abbreviations, ensuring compliance with Australian billing conventions ("Attending upon" rather than "Meeting with"), and adding specificity where the initial generation was vague.
Polish is completely free. No credit deduction, no usage limit. You can polish the same entry multiple times if the first polish does not quite hit the mark. "Polish All" applies polish to every entry in the batch with a single click.
Polish is particularly effective on entries generated from brief or unclear audio. If the original recording was noisy or the speaker was mumbling, the initial generation may produce vague descriptions. Polish takes those vague descriptions and infers more specific language from the available context.
Tip 3: Thorough vs Conservative — Know When to Switch
The Thorough/Conservative toggle controls which entries are visible, but most users leave it on the default (Thorough) and never experiment. Here is when each mode shines.
Use Conservative when: You are processing a straightforward matter and want a quick, reliable export. Conservative hides medium and low confidence entries, showing only the entries the AI is most certain about. For a clear, well-recorded 30-minute meeting, Conservative typically captures 90%+ of the billable activities and the export is essentially ready to import without editing.
Use Thorough when: The source material is complex — a multi-issue conference, a lengthy document review, or a recording with multiple speakers discussing different matters. Thorough shows every entry the AI generated, including speculative ones marked as low confidence. You review each one and delete or keep. This catches billable activities that Conservative would hide — the 5-minute discussion about a related matter, the brief instructions about a disbursement, the aside about the next hearing date.
Switching between modes does not use credits and does not trigger a new API call. It is purely a front-end filter on the already-generated entries.
Tip 4: Multi-File Document Upload Strategy
When using the Docs & Emails tab, the order and grouping of your file uploads affects the quality of the output. The AI processes files in context — if you upload five related emails from the same matter, the AI recognises the thread and generates entries that reflect the progression of the correspondence rather than treating each email in isolation.
For large documents (50+ pages), the AI uses automatic chunking — splitting the document into manageable segments, processing each, and merging the results. You do not need to split documents manually. However, if you are processing a very large brief (200+ pages), consider uploading it in logical sections (pleadings, evidence, submissions) to get more targeted billing descriptions for each section.
Tip 5: Voice FAB for Quick Capture
The floating action button (FAB) in the bottom corner of the screen is a quick voice recording tool. Tap it, dictate a brief note about a billable activity, and it auto-submits for processing. This is designed for the moments between activities — after a phone call, walking back from court, or between client meetings.
The FAB workflow is: tap → speak → stop → auto-submit. There is no file management, no upload step, no button clicking. It is the fastest path from "I just did something billable" to "there is a time entry for it."
For lawyers who struggle with end-of-day time reconstruction, the FAB is a game changer. Instead of trying to remember at 6pm what you did at 10am, capture each activity in real time with a 15-second voice note. At the end of the day, your entries are already generated and waiting for review.
Tip 6: Clio Push Best Practices
If you have connected your Clio account, the Push to Clio feature sends entries directly to Clio as activities. Here are the practices that make this work smoothly.
Matter number matching: LexUnits matches entries to Clio matters using the matter number in the Session Context or in each entry's associated_matter field. The matching is automatic — LexUnits searches your Clio matters for a display number match. For reliable matching, use the exact Clio display number (e.g., "MAT-2024-0532") rather than informal matter references. If the matter number does not match any Clio matter, the push for that entry will fail — check the entry's matter field before pushing.
Rate matching: When you push to Clio, LexUnits sends the rate associated with the entry. If the rate does not match the fee earner's rate in Clio, the Clio rate takes precedence for invoicing purposes. For consistency, ensure your LexUnits rate list matches your Clio rate configuration.
Review before pushing: Once pushed, entries appear in Clio as activities. They can be edited in Clio, but it is easier to get them right in LexUnits first. Polish, review, confirm matter numbers, then push.
Tip 7: Meeting Minutes as a Billing Multiplier
The Meeting Minutes tool (accessible from the sidebar or the AI Tools cards) is often used as a standalone feature — generating minutes from a meeting recording. But it is also a billing multiplier when used alongside the core billing workflow.
The workflow: upload a recording on the Record & Bill tab to generate billing entries (1 credit for transcription). Then use the same transcript (which was already generated) to create Meeting Minutes — this second step is free. You now have both billing entries and professional meeting minutes from a single upload and a single credit.
This is particularly valuable for client-facing meetings where you need to send minutes to the client and file billing entries with your firm. One recording, two outputs, one credit.
Tip 8: Attendance vs Follow-Up Tags
LexUnits classifies each billing entry as either an "Attendance" entry (work performed during the meeting — listening, discussing, advising) or a "Follow-Up" entry (work that needs to be done after the meeting — drafting documents, sending correspondence, conducting research).
This classification is useful for two purposes. First, it helps with work management — the follow-up entries are effectively a task list generated automatically from your meeting. Second, it affects billing timing. Attendance entries are billed as of the meeting date. Follow-up entries should be billed as of the date the work is actually performed, which may be days or weeks later. Review the follow-up entries and adjust the date when the work is completed.
Tip 9: Optimise Your Audio Quality
Recording quality directly affects billing entry quality. You do not need professional equipment — a smartphone placed on the table works well for most meeting rooms. But a few practices make a noticeable difference.
State names at the beginning of the recording: "This is [Lawyer Name] in conference with [Client Name] regarding matter [Number]." The AI uses this information to generate accurate billing descriptions that reference the correct names and matter.
For phone calls, use your phone's voice recorder app on speaker phone, or use a Zoom/Teams recording. The key is capturing both sides of the conversation — a recording that only captures your voice will generate entries based only on what you said, missing the context from the other party's statements.
Avoid background noise where possible. A quiet meeting room produces dramatically better transcription than a busy café. If you must record in a noisy environment, speak clearly and slightly louder than normal.
Tip 10: Use Document Summary Before Full Review
When you receive a large document for review — an expert report, a lengthy affidavit, a contract — use the Document Summary tool first (1 credit). Read the AI summary to understand the document's structure, key points, and potential issues. Then conduct your full review with targeted attention to the sections the summary flagged.
After your review, use the Docs & Emails tab to generate billing entries for the review work (1 credit). You have now used 2 credits total and have both a structured summary and detailed billing entries — work that would typically take 15–20 minutes of manual writing.
Ready to Work Faster?
Sign up free with 10 credits and try these workflows yourself. No credit card required.
Start Free TrialIs AI Polish free in LexUnits?
Yes. AI Polish — both individual entry polish (the sparkle icon) and Polish All (batch polish) — is completely free with no credit deduction. You can polish entries as many times as you like. This is intentional: polishing is a refinement step that should never discourage you from improving your billing descriptions.
What is the difference between Thorough and Conservative mode?
Thorough mode shows all generated billing entries regardless of confidence level. Conservative mode hides medium and low confidence entries, showing only high-confidence entries. Both modes are front-end filters only — switching between them does not use credits or trigger a new API call. Use Conservative for quick, reliable exports; use Thorough to review every potential entry and decide which to keep.
How do I get better results from audio recordings?
Three things improve audio-to-billing quality significantly: fill in the Session Context before generating (matter number, date, timekeeper, client name), use a decent recording — even a phone placed on the table works if the room is quiet, and for multi-speaker meetings state names at the beginning ("This is [Name] meeting with [Client Name]"). The AI uses these signals to generate more accurate billing descriptions.
Last verified: April 2026.