Process

  • Self-directed screens. A public-company target came out of an in-house screen: profile, model, a 40-question DD list, tracker, overview deck and cap-table analysis, reaching an LOI in fifteen days. The comparable set was built by hand from national pharmacy and healthcare accreditation registries (89 companies) rather than pulled from a database.
  • Cold outreach off public signals. A founder posted that his company was winding down; outreach went out the same day. When the deal was killed internally, the relationship stayed warm.

Select valuation engagements.

EngagementPurposeDelivered
Post-close valuation of an acquired public-company subsidiary Transaction Owned end-to-end through review: comparable universe, 10-year DCF, sensitivities and the board deck
Public-company target waterfall LOI support First-pass distribution waterfall supporting the offer structure
Distressed pro-forma and distribution waterfall Self-initiated Valued on how much lenders were likely to recover, rather than on a headline sale price
Merger valuation, three-party structure Board Four alternative valuation framings, sum-of-parts
Comparable-companies valuation, major retail partner Board / principal Full trading-comparables set built from primary filings. The peer universe was screened and defended, each name normalized for capital structure and one-off items, benchmarked on revenue and EBITDA multiples with the range, median and implied value shown rather than a single point. Requested and delivered the same day.

Counterparties withheld.

Primary sources, always.

State regulatory merger filings, foreign company-registry returns cited at filing-reference level, federal contract databases, FDA prescribing information, court dockets, DOIs.

  • Read it where it was actually documented. A national pharmacy’s mandatory-enrollment mechanic was reconstructed from a school district’s benefits portal, a place nobody thinks to look. Reading it there settled a question the marketing material left open.
  • Read the opinion, not the summary. On another diligence, a claim made in a management presentation turned on how a court had actually ruled. Reading the opinion changed the answer, and the memo was written from it.

Fieldwork beats the data room

Dealers were called directly to test a claimed commercial relationship. What came back did not match the paperwork. Read onClose

The traction thesis was rebuilt around what could be confirmed by the people actually selling the product.

Audit trails are built into the models, not bolted on

Every reporting model carries an independent recomputation of its own totals. Read onClose
  • Computed twice, by different routes. Figures are rebuilt from row-level detail along a separate path from the summary rollup, compared against it, and held to a stated tolerance.
  • A break surfaces as a variance. A broken rollup shows up as a visible difference rather than as a clean-looking number.
  • The check travels with the file. It runs every time the model does, including the times nobody thought to look, and a reviewer inherits the test rather than having to trust the author.

It sits inside the model rather than in a review step because rollup errors are silent by construction: a mis-scoped sum returns a plausible number, not an error, and it will survive any amount of reading. The only reliable way to catch one is to force two independent computations to agree.

Standing intelligence products

A GLP-1 competitor deck maintained since December 2025, plus two weekly products. Read onClose
  • GLP-1 competitor intelligence. A 25+ competitor deck, maintained continuously since December 2025.
  • Weekly financial-markets brief to the executive list.
  • Weekly clippings product to the internal intelligence distribution.

Contact

Solomon Dworsky · Miami, Florida

The pages here stop at what was built and what it produced. If you want the detail underneath any of it, ask.

Private equity transaction work, healthcare technology, and applied AI in the investment process.