Key Deposition Excerpts

Critical testimony from Priya Sharma, Maya Chen, and Daniel Osito.

⚠️ Highly Confidential

Deposition transcripts are among the most sensitive litigation documents. They contain client testimony, opposing party admissions, and case strategy. Extreme care is required before using any tool with this content.

Tool Selection Opportunity

Task: Identify inconsistencies between witnesses, find impeachment material, extract key admissions

Non-AI Options
  • Manual comparison across transcripts with annotation
  • Spreadsheet tracking key statements by witness and topic
  • Index of testimony by subject matter
AI Options (Tools You Have)
  • CoCounsel/Protégé: Deposition analysis within privileged platforms—the appropriate choice if using AI
  • Gemini/Copilot: Uploading deposition transcripts to general-purpose AI would expose highly sensitive client information
Selection Considerations
  • Deposition transcripts contain sworn testimony with strategic significance—manual review may be warranted regardless of AI availability
  • The highlighted portions contain potential impeachment material—compare to interrogatory responses
  • For trial preparation, understanding the nuance of testimony often requires human judgment
  • This may be a case where non-AI analysis is clearly preferable

Deposition Excerpts

CAUSE NO. 2026-CV-04521


Deposition of Priya Sharma

Lead Developer, Vertex Solutions LLC | Taken: March 10, 2026

[Pages 34-38: Testing Procedures]

Q: Ms. Sharma, you were the lead developer on the GreenLeaf project, correct?

A: Yes, that's correct.

Q: Can you describe the testing you performed before delivering the software?

A: Sure. We did our standard testing protocol—unit tests for individual components, integration testing to make sure the modules worked together, and then some user acceptance testing with sample data.

Q: Did you test the inventory tracking function specifically for accuracy?

A: Yes, we ran tests where we entered known quantities and verified they displayed correctly.

Q: What quantities did you use in your testing?

A: We used smaller datasets. Maybe 10 to 50 items at a time. Standard test scenarios.

Q: Did you ever test with quantities of 100 or more?

A: I don't recall testing with quantities that large specifically. Our test cases focused on verifying the functionality worked, not stress testing with large volumes.

[Pages 45-52: Reporting Function]

Q: Let's talk about the reporting function. Did you design a maximum capacity for reports?

A: The system was designed for typical business reporting. We didn't set a hard limit, but there are practical constraints based on browser memory and server resources.

Q: Were you aware that GreenLeaf had over 2,000 SKUs?

A: I knew they were a mid-sized operation, but I don't recall the exact number being discussed during development.

Q: Did you test the reporting function with 2,000 line items?

A: No. We tested with smaller datasets—a few hundred items.

Q: So you never tested whether the system could generate reports for GreenLeaf's actual inventory?

A: We tested the reporting functionality. We verified it could generate reports. The spec didn't require any specific capacity.

[Pages 67-73: "Acceptable Industry Tolerances"]

Q: The specification says the system should "maintain accuracy within acceptable industry tolerances." What does that phrase mean?

A: It means the system should be accurate within reasonable expectations. No software is perfect—there can be minor variances due to timing issues or rounding.

Q: What variance would be acceptable?

A: For inventory systems, you'd expect high accuracy. Maybe 1-2% variance in some circumstances.

Q: Would a 15-20% variance be within "acceptable industry tolerances"?

A: No, that would be outside what I'd consider acceptable. That's a significant discrepancy.


Deposition of Maya Chen

CEO, GreenLeaf Technologies, Inc. | Taken: March 12, 2026

[Pages 28-35: Contract Negotiations]

Q: Ms. Chen, who negotiated the software contract on GreenLeaf's behalf?

A: I did, primarily. With input from our operations manager, James Morrison.

Q: During negotiations, did you discuss the size of GreenLeaf's inventory?

A: Yes. We told Vertex we had about 2,000 SKUs and were growing. We needed a system that could scale with us.

Q: Did you discuss specific accuracy requirements?

A: We discussed that we needed accurate inventory tracking. That's the whole point of inventory software. Daniel assured us their system would handle our needs.

[Pages 48-55: Impact of Defects]

Q: Can you describe how the software problems have affected GreenLeaf's operations?

A: They've been significant. We invested in this software to replace our manual spreadsheet tracking, which was becoming unmanageable as we grew. Instead, we've had to keep using the spreadsheets because we can't trust the software.

Q: Have there been any specific incidents caused by the inventory inaccuracies?

A: Yes. In February, we nearly ran out of our EP-500 containers because the system showed we had plenty in stock. We caught it during a manual check, but if we hadn't, we would have missed a major customer order.


Deposition of Daniel Osito

Managing Partner, Vertex Solutions LLC | Taken: March 15, 2026

[Pages 41-47: Defect Response]

Q: When GreenLeaf first complained about defects, what did you do?

A: I asked our lead developer to investigate. She looked into it and reported back that the system was working as designed.

Q: After receiving the formal defect notice on January 27th, what investigation did Vertex conduct?

A: Priya reviewed their documentation. She concluded the issues were either user error or requests for functionality beyond the contract scope.

Q: Did anyone visit GreenLeaf's facility to observe the problems firsthand?

A: No, we didn't think that was necessary.

Q: During the 15-day cure period, did Vertex make any changes to the software?

A: We didn't believe there were defects that needed curing. The software met the contract specs.

Q: So Vertex made no attempt to fix the issues GreenLeaf reported?

A: We didn't view them as valid defect claims. They were asking for things beyond what we agreed to deliver.