The draft presentence report puts the loss at $11.4 million. Forty-seven investors, nine bank accounts, three entities, four years. Your client kept the books for the last 19 months of it and had signature authority on 2 of the 9 accounts. She pleaded guilty to one count of wire fraud conspiracy under 18 U.S.C. § 1349, and probation has attributed all $11.4 million to her.
On November 1, 2026, that number starts to matter in a new way. The Sentencing Commission’s inflation adjustment to the §2B1.1 loss table takes effect that day unless Congress acts. It is the first update to the table since 2015. The threshold for a 14-level increase moves from $550,000 to $750,000. The threshold for a 20-level increase moves from $9.5 million to $15 million. Courts generally apply the Guidelines Manual in effect on the day of sentencing.
So the question that sets her advisory range is not how much the scheme took. It is how much of it is hers.
Two numbers, six levels apart
Under §1B1.3(a)(1)(B), a defendant in a jointly undertaken scheme answers for the acts of others only when those acts were within the scope of what she agreed to, in furtherance of it, and reasonably foreseeable. The commentary is direct about timing. Relevant conduct does not include what other members of a conspiracy did before the defendant joined, even if she later learned of it.
Suppose the records show that investor money reaching her two accounts, after the day she started, totals $640,000. Under the amended table, $11.4 million is an 18-level increase. $640,000 is a 12-level increase.
With a base offense level of 7 and every other adjustment held equal, that is level 25 against level 19. In Criminal History Category I, level 25 carries an advisory range of 57 to 71 months. Level 19 carries 30 to 37.
The government has to support its loss figure. The defense has to show the court a better one. Both sides are working from the same production.
What the production actually contains
- Bank records for nine accounts at three institutions, produced as spreadsheets with tens of thousands of rows
- The case agent’s loss schedule, with one row per investor and one total at the bottom
- The investor ledger, with names, dates, and amounts
- Account opening records and signature cards showing who had authority over which account, and when
- Payroll records showing her hire date and role
- Toll records and call detail records for the principals’ phones
- Secretary of State filings for the three entities
Every one of those sources names people, accounts, dates, and amounts. None of them, read alone, says which dollars passed through her hands.
Where the tabs stop working
The usual approach is to rebuild the agent’s schedule in a workbook. One tab per account. A tab for investors. A tab for entities.
Then the questions start. Which investor deposits landed in her two accounts, and which landed in the seven she never touched? Of the money that reached her accounts, how much arrived before her start date? When funds moved from a principal’s account into one of hers and back out to an earlier investor, whose transaction is that?
Each answer is a lookup across three tabs and two date columns, run by hand, forty-seven times. Change one assumption, and you run them all again.
In SentVi, the same records are imported as entities and relationships. Every investor, account, company, and person is an entity. Every deposit, transfer, signature authority, employment date, and call is a relationship with an attached date and amount. The scheme draws itself as one network, and her place in it is visible: two accounts on one side of the chart, connected to everything else through the two principals.
The three views that matter at sentencing
The timeline. Set the time-range slider to open on her first day. Everything before it drops off the chart: twenty-nine of the forty-seven investors and twenty-nine months of deposits. What remains is the scheme as it existed while she was in it. That is the picture the relevant conduct commentary describes, and it is an exhibit.
The network. Social network analysis ranks every entity by how connected it is and by whether it sits between otherwise separate groups. In this chart, the two principals score highest on both. They open the accounts, take the investor calls, and sit between the investors and the money. The bookkeeper sits at the edge, attached to two accounts and one supervisor. Role adjustments under §3B1.1 and §3B1.2 turn on decision-making authority, participation in planning, and the nature and extent of what the defendant did. A chart does not decide those questions. It shows the court the structure they ask about.
The path. Select an investor and one of her accounts and run Shortest Path or All Paths between them. Either the money gets there, and the chart shows each hop it took, or there is no path at all. Run across the investor list; that sorts the loss schedule into deposits that reached her accounts and deposits that never did. The dollar figure in the objection still comes from the bank records. The chart shows which records belong in the count.
The same chart cuts both ways
Prosecutors and agents run this analysis too, and the data does not care who is asking. If her phone appears in the call records ahead of each investor wire, or her accounts turn out to be the bridge between two groups of investors, the chart shows that as well. A defense team is better off seeing it before the sentencing memorandum is filed than in the government’s response to it.
Scope and role are questions the records can answer. Neither side has to argue them from a spreadsheet total.
From the chart to the sentencing memorandum
Objections to the presentence report, the sentencing memorandum, and the hearing all require the same analysis, presented in a form the court can read in a minute. SentVi exports network diagrams and timelines as exhibits anyone can read without training. SentVi Reader lets co-counsel and a forensic accountant open the interactive chart without each needing a license.
Because SentVi runs on your own hardware and can operate fully disconnected from the internet, discovery produced under a protective order stays on machines you control.
The question the defense should ask
Not “how big was the fraud,” but “which part of the network is hers, and since when.” A spreadsheet gives you the government’s total. Link analysis shows you what is inside it.
See the workflow in the video below, or request a demo: https://sentvi.com/link-analysis-book-a-demo/
