Paralegal Deposition Review: A Tagging System

A 150-page deposition transcript is a raw material. The attorney needs specific things from it: the witness's account of dates and events, contradictions with other testimony, admissions that support or undermine claims, foundation for cross-examination. Without a system for finding these passages, the attorney reads the full transcript before every meeting with it.

A tagging system solves this by creating a searchable index during first-pass review. Paralegals apply it once; attorneys retrieve by category from then on.

BrassTranscripts produces speaker-labeled, timestamped transcripts from deposition audio in minutes. The transcript becomes the raw material. The tagging system is what makes it case-ready.

Quick Navigation

• Why Tagging Works Better Than Highlighting
• Standard Tag Categories for Deposition Review
• Three Tagging Methods
• The Word Comment Tag Method
• The Spreadsheet Index Method
• Using AI for Pre-Tagging
• Delivering the Tagged Transcript to Attorneys
• Frequently Asked Questions

Why Tagging Works Better Than Highlighting

Highlighted transcripts fail for the same reason color-coded paper notes fail: retrieving the information requires a full visual scan. To find the witness's contradictory statement about when the meeting occurred, you read through every yellow highlighted passage until you reach it.

Tagged transcripts work differently. A tag is a label attached to a passage: CONTRADICTION, ADMISSION, TIMELINE, KEY FACT. To find contradictions, filter for CONTRADICTION. To find the witness's account of the meeting date, search for TIMELINE entries mentioning the meeting.

The tagging pass happens once, during first review. After that, the transcript works as an indexed reference document rather than a long text file.

BrassTranscripts transcripts include timestamps for every speaker turn, which makes two common tasks significantly faster:

Locating the original audio. Tag entry "CONTRADICTION — Speaker 2 states meeting occurred on Tuesday (04:22:15)" lets you jump directly to that moment in the audio recording if you need to hear the original testimony.

Citing testimony precisely. Brief citations with timestamp or page number are easier to generate from a timestamped transcript than from a page-only reference.

Standard Tag Categories for Deposition Review

Tag categories should be consistent across a case's transcripts so the same labels work for all depositions. Recommended standard categories:

TIMELINE — any statement about dates, times, sequences, or the order of events. These build the factual chronology.

CONTRADICTION — any statement that conflicts with other testimony, the witness's own prior statements, or documentary evidence. Flag with a brief note about what it contradicts.

ADMISSION — statements favorable to the client's position. Admissions about the defendant's knowledge, the condition of property, what occurred, who was responsible.

DENIAL — explicit denials of key facts. These are equally important: if the witness later reverses, the original denial is impeachment material.

CHARACTER/CREDIBILITY — testimony that affects how the jury will evaluate this witness: prior inconsistent statements, demonstrated uncertainty, evasive answers, statements about what the witness did or didn't know.

FOUNDATION — testimony that establishes or undermines the foundation for evidence, expert testimony, or procedural claims.

FOLLOW-UP — passages that require additional investigation, document production, or questions for the next deposition.

KEY FACT — material facts that don't fit the above categories but are likely to appear in the case summary, brief, or jury instructions.

Adapt these categories to the case type. A personal injury case may add INJURY DESCRIPTION and PRIOR CONDITION. A contract dispute may add INTERPRETATION and COURSE OF CONDUCT. Keep the category list to 6-10 labels — more creates tagging ambiguity and slows review.

Three Tagging Methods

Method 1: Word comments. Open the TXT transcript in Word or Google Docs. As you read, insert a comment at each taggable passage with the category label. Comments are searchable (Command+F on Mac, Ctrl+F on Windows), appear in the comment panel for visual review, and survive when the document is shared.

Method 2: Spreadsheet index. Build a spreadsheet with columns for Page/Timestamp, Category, Speaker, Quote, and Notes. As you read, add a row for each tagged passage. The result is a case-specific database: sort by Category to group all CONTRADICTION entries, or filter by Speaker to see everything the opposing expert said about a specific topic.

Method 3: AI-assisted pre-tagging. Paste the full transcript into an AI tool with a tagging prompt. The AI identifies and pre-tags passages by category. The paralegal reviews the AI output for accuracy, catches misses, and resolves ambiguous categorizations. This method works best for straightforward factual testimony; it requires more verification for credibility and foundation questions that require legal judgment.

Each method works. The right choice depends on how the attorney uses the output and what tools the firm has available.

The Word Comment Tag Method

This method requires only Microsoft Word or Google Docs and produces a shareable, annotated document.

Step 1: Download the TXT transcript from BrassTranscripts and paste it into a blank Word document.

Step 2: Read through the transcript. When you reach a taggable passage, select the relevant text, right-click, and insert a comment. Type the category label first, then a brief descriptive note: CONTRADICTION — Witness states he was not at the facility on March 12, but see Exhibit 4 showing his badge swipe at 9:47am.

Step 3: After tagging, use Word's comment panel (Review → Show Comments) to see all tags listed sequentially. This is the attorney's quick-reference view.

Step 4: To create a category-specific view, search comments in the panel for the tag label. All CONTRADICTION comments become visible together.

Advantage: Familiar tools, no additional software, comments survive document sharing.

Limitation: Searching the comment text rather than filtering rows is slightly slower than a spreadsheet for large transcripts.

The Spreadsheet Index Method

For cases with multiple depositions, a spreadsheet index scales better than annotated documents. All deposition testimony lives in one searchable table.

Structure:

Step 1: Create the spreadsheet with the columns above.

Step 2: Work through the transcript, adding one row per tagged passage. The Timestamp column comes from the BrassTranscripts transcript (each speaker turn is timestamped).

Step 3: As additional depositions are processed, add rows from each deposition to the same spreadsheet.

Step 4: Filter by Category to pull all CONTRADICTION rows across all depositions. Filter by Witness to see everything a specific witness said in a specific category.

This scales to 20-30 depositions in a complex matter. The attorney retrieves cross-deposition contradictions with a filter; no manual page-scanning across stacks of transcripts.

The JSON format from BrassTranscripts, which includes structured speaker and timestamp data per utterance, imports directly into Excel or Google Sheets for firms that want to pre-populate timestamp and speaker columns programmatically.

Using AI for Pre-Tagging

After BrassTranscripts produces the transcript, AI tools can perform a first-pass tagging that the paralegal then verifies. The prompt structure:

Below is a deposition transcript from [witness name], taken in [case description]. 

Review the full transcript and identify passages that fall into these categories:
- TIMELINE: Any statement about dates, times, sequences, or the order of events
- CONTRADICTION: Any statement inconsistent with other testimony or evidence (flag the inconsistency)
- ADMISSION: Statements favorable to plaintiff's position
- KEY FACT: Material facts about [specific issue in dispute]
- FOLLOW-UP: Passages requiring investigation or follow-up questions

For each tagged passage, provide: Timestamp (if available), Category, Speaker, exact quote (up to 2 sentences), and a one-sentence note about why it's tagged.

[paste full transcript]

The AI output is a structured list of tagged passages. The paralegal's job becomes verification: Are the CONTRADICTION tags accurate? Were any ADMISSIONS missed? Is the TIMELINE complete?

This approach is most useful for long transcripts (3+ hours) where first-pass paralegal review is time-consuming. The AI pre-tagging covers the straightforward identifications; the paralegal focuses review time on passages requiring legal judgment.

For specialized deposition analysis prompts — contradiction detection, timeline extraction, cross-examination preparation — see the legal professional AI toolkit.

Delivering the Tagged Transcript to Attorneys

The attorney's version of the tagged transcript is the output of paralegal review, not the raw annotated document. Standard deliverables:

Deposition summary memo. A 2-5 page structured summary organized by tag category. TIMELINE section lists key dates and events in chronological order with transcript citations. CONTRADICTION section lists each contradiction with the conflicting evidence. ADMISSIONS section lists favorable testimony with quotes and timestamps. FOLLOW-UP section is an action list.

Annotated transcript. The full transcript with comment tags visible. Useful when the attorney needs to read specific sections in context rather than from summary quotes.

Spreadsheet index (for multi-deposition matters). The cross-deposition index showing all tagged passages sorted by category and witness.

For deposition audio that hasn't been transcribed yet, BrassTranscripts processes the file and returns a speaker-labeled transcript in minutes. For batches of case deposition recordings, the bulk service processes all files concurrently with no minimum count.

The legal practice AI tools guide covers the full workflow for solo and small firm case preparation.

Frequently Asked Questions

How Does Tagging a Deposition Transcript Differ from Highlighting?

Highlighting marks text visually; tagging assigns searchable category labels. A highlighted deposition requires visual scanning to find relevant passages. A tagged deposition can be filtered: show all passages tagged CONTRADICTION, or all passages tagged TIMELINE. Tagging is typically done in a spreadsheet or document management system with a tag column, or using comment tags in Word or Google Docs. The result is a structured index that attorneys can search by topic rather than page number.

What Deposition Transcript Format Works Best for Paralegal Review?

TXT for manual review and annotation in a word processor; JSON for any workflow that involves structured data processing, spreadsheet import, or case management software integration. The JSON transcript from BrassTranscripts includes speaker labels and timestamps for each utterance, which is useful for building a row-based tagging spreadsheet. Download both formats and use whichever fits the review tool.

How Long Does It Take to Tag a Two-Hour Deposition Transcript?

Tagging time depends on how many categories you're tracking and how densely the testimony covers those categories. For a standard deposition covering 4-6 topic areas, first-pass tagging of a two-hour transcript — assigning category labels to relevant passages as you read — typically takes 60-90 minutes. A second pass to verify and refine tags adds 20-30 minutes. Building the index or summary from tags takes an additional 15-30 minutes. Total: 2-3 hours for a full two-hour deposition, compared to 4-6 hours for a first-pass manual read without a structured tagging system.

Can AI Tools Help with the Tagging Process After Transcription?

Yes. After BrassTranscripts produces the transcript, an AI tool (ChatGPT, Claude, Gemini) can process the full text and pre-tag passages by category. The prompt specifies the categories: contradictions, admissions, timeline events, key facts, character credibility. The AI output is a structured list of tagged passages that the paralegal then reviews for accuracy and completeness. This reduces first-pass tagging time significantly and focuses paralegal review on verification rather than identification.