Source the next deal, read every document, report to committee.
You work the whole deal off one model: sourcing against your thesis, the market mapped, every CIM read into proprietary data with its provenance, and the committee memo built from the record. Built around your firm's edge, and run for you.
Surface the targets before a banker does.
You start from your thesis. A seed expands into a scored, validated collection across millions of companies, weighed against your mandate and your prior pipeline. The targets you should be seeing come forward before a banker brings them.
Every target is matched against the companies already on your record, so the same business never shows up twice and a name you passed on two years ago surfaces with its history attached. You start the week on a ranked, validated list instead of a banker's teaser.
Grow Light Manufacturers
Sourced from your thesis, scored and validated, deduped to one record each.
| # | Tier | Company | Keyword | AI relevance | Source | Enrichment |
|---|---|---|---|---|---|---|
| 1 | T7 | GrowTech Solutions growtechsol.com | 85.34 | 72 likely | LinkedIn Search | enriched |
| 2 | T6 | Lumina Grow Systems luminagrow.com | 81.0 | 68 likely | Web Search | scraped |
| 3 | T4 | AquaLux Inc aqualux.io → techwater-corp.com | 62.1 | 45 unlikely | CSV upload | redirected |
| 4 | T8 | Photon Ag photonag.com | 88.9 | 79 likely | Referral | validated |
Make every target comparable across the pipeline.
Prophet reads each company's products into a normalized capability taxonomy with clusters, so each target becomes a structured position you can set against every other company you have looked at.
Maturity and whitespace fall out of the same structure: where a target is deep, where it is thin, and where the market is open.
Capability clusters · Aviation software
The same company, resolved across your pipeline, your diligence, and the market.
The record resolves continuously, so when a company comes back around it is already there, not rebuilt by hand: the relationship history, the ownership, the people, and every prior touch held on one company. Your team works it from one record instead of three half-versions.
Sourcing, the market position, and the deal history all hang off that record, so when a company comes back around you already know it.
Northwind Field Software
Field-service management software for HVAC & mechanical contractors
Company basics
- Employees
- 178 (+19%)
- HQ
- Austin, TX
- Founded
- 2014
- Leadership
- Founder-CEO, 11 yr
Funding & ownership
- Raised
- $32.0M
- Last round
- Series B
- Lead
- Summit Partners
- Status
- Inbound
Financials
- Revenue
- $24.1Mp.14 ↗
- ARR
- $21.6M
- Growth
- +28%
- EBITDA
- 14%
Market position
- Segment
- Mid-market FS
- Rule-of-40
- 42 ✓
- Breadth
- 14 / 18
- Fit
- Strong
Products & capabilities — decomposed by Prophet
Schedules technicians, assigns jobs, and runs the day in the field.
Turns completed work into invoices and keeps recurring agreements current.
Tracks installed equipment and the certifications the work depends on.
Every CIM you read becomes proprietary data that compounds.
AI reads every CIM and data-room file into structured records: terms, figures, and segments, each one traced to the page and region it came from. Because it extracts onto your resolved record, your analyst verifies against the source in one click instead of retyping it.
Across every deal you look at, that structured data accumulates into a proprietary dataset only you have, queryable and current. The more you diligence, the sharper the edge you are building.
Acme Corp Acquisition
Playbook · Financial Diligence
Deal score
GOFinancials
FY 2021–2023| Metric | 2021 | 2022 | 2023 |
|---|---|---|---|
| Revenue | $2.5M | $3.1M | $3.8M |
| EBITDA | $400k | $550k | $725k |
| EBITDA margin | 16.0% | 17.7% | 19.1% |
| Gross margin | 61% | 63% | 64% |
| Net retention | 88% | 90% | 92% |
Tie every figure in the CIM back to the document that proves it.
The model reconciles the figures management presents against the underlying data room: bank statements, the GL, AR aging. Each revenue figure and each adjusted-EBITDA addback ties back to the exact source doc and page, so your team verifies the story instead of taking it on faith.
Where the numbers do not agree, an agent drafts the variance and the document it should reconcile to, and holds it for your analyst to clear before the memo. Because it reads off data that agrees, the draft is one your team can trust. An owner-comp addback presented at $0.62M that the bank statements support at $0.41M is flagged, not buried: the $0.21M carries straight into adjusted EBITDA.
CIM vs. data room
Every figure traced to the underlying document, with the variance held on the row.
| Figure | CIM | Resolved | Variance | Status | Traced to |
|---|---|---|---|---|---|
| Revenue (FY23) | $42.5M | $42.1M | −$0.4M | MISMATCH | gl_2023.xlsx |
| Reported EBITDA | $8.10M | $8.10M | $0 | ✓Ties | gl_2023.xlsx |
| Owner comp addback | $0.62M | $0.41M | −$0.21M | MISMATCH | stmts Q1–Q4 |
| One-time legal addback | $0.34M | $0.34M | $0 | ✓Ties | gl_2023.xlsx |
| Non-recurring rent | $0.18M | $0.18M | $0 | ✓Ties | stmts Q1–Q4 |
| Adjusted EBITDA | $9.24M | $9.03M | −$0.21M | MISMATCH | gl_2023.xlsx |
Owner-comp addback presented at $0.62M; bank statements support $0.41M. The $0.21M gap carries straight into adjusted EBITDA. Verify against stmts Q1–Q4 before the memo.
Report to committee off the same record you diligenced.
An agent drafts the IC memo and the quality-of-earnings summary from the structured record, and every figure on the slide links back to the CIM page and the model cell it came from. It drafts; your deal team approves and takes it to committee. Because it draws on the resolved record and the market map you built, the position you take is the one your data supports.
The same record carries through the hold and into the exit: portfolio performance on one model and the buyer universe, current from first look through the sale.
Investment committee memo
Revenue compounded at 26% to $42.5M TTM, with EBITDA margin expanding 1.4 points to 19.1% as the cost base scaled sub-linearly with bookings.
Hold every company to the plan that won the deal.
Through the hold, each portfolio company is measured against its own underwriting: revenue and EBITDA, plan versus actual, by company. The model reads each company's systems, holds the KPIs current, and flags the holdings drifting off plan before the quarter closes.
When a company runs behind, an agent drafts the variance against plan and the metric that moved, and routes it to the deal partner for the next operating review under your permissions. A holding tracking EBITDA at 77% of plan because gross margin slipped is on the agenda, not a surprise at year-end.
Actuals vs. underwriting plan
Each holding measured against the plan that won the deal, current from company systems.
| Company | Revenue · plan / actual | EBITDA · plan / actual | Status | Read from |
|---|---|---|---|---|
Northwind Logistics vintage 2022 | $60.4M104% plan $58.0M | $12.3M106% plan $11.6M | On plan | NetSuite |
Tessera Health vintage 2023 | $41.7M102% plan $41.0M | $7.4M95% plan $7.8M | Watch | Sage Intacct |
Cobalt Software vintage 2021 | $30.1M90% plan $33.5M | $7.2M77% plan $9.4M | Behind plan | QuickBooks |
Harbor Foods vintage 2024 | $24.3M93% plan $26.0M | $2.5M81% plan $3.1M | Behind plan | NetSuite |
Cobalt Software is tracking EBITDA at 77% of plan: revenue held but gross margin slipped 3 points since close. Flagged for the next operating review with the deal partner.
See it on your own deal stack.
Tell us how your firm sources and reads deals today. We will show you the model built around your edge, run for you.