What you can do in Aqmen
One workspace where market models, competitive grids, and company financials stay structured, source-cited, and scenario-ready. You build them the way you always have, with AI agents taking the grunt work off your plate.
01 · Market Sizing
Markets as driver trees, not spreadsheets
Structure TAM, SAM, and SOM as a value driver tree (the framework top firms use) with dimensions, variables, and time series you can interrogate. Split a segment, swap a dimension, re-cut the market: the model restructures instead of breaking.
Historicals are asserted with sources. Forecasts are driven by explicit hypotheses. Nothing in the tree is a typed-in number nobody can defend.
market driver tree · 2025
price × volume · 14 segments
02 · Competitive Landscape
Every player on the same grid
Benchmark the field on a standardized dataset of archetypes, purchase criteria, and positioning, so players are compared on the same dimensions instead of a patchwork of profiles.
Add custom fields for what makes this market different, and keep every entry backed by its source.
competitive landscape · 8 players
| player | archetype | revenue | nps |
|---|---|---|---|
| Alpha Corp | Integrator | €120M | 44 |
| Betaworks | Specialist | €85M | 61 |
| Gamma Sys | Challenger | €40M | 52 |
03 · Company Analysis
The target's P&L, modeled and traceable
Build a full financial model of the target from revenue drivers down to EBITDA, with forecasts driven by the same hypothesis-and-scenario engine.
Valuation sits on top of the model, not beside it: change an assumption and watch it flow through to the number that matters.
target P&L · FY2025
EBITDA margin 18.3% · computed
Agent-first
Or skip the clicking, ask.
Build a bottom-up market model for European industrial software.
Benchmark the top 8 players on pricing model and NPS.
Apply a downside scenario to the 2027 plan. What changes?
Your AI agent connects straight to the project over MCP. Everything it does lands as real artifacts, trees, grids, models, cited, structured, and waiting for your review.
04 · Projects & Collaboration
The whole engagement, one shared project
A project is the engagement: the market sizing, the competitive landscape, the company model, and every source behind them, in one place. Invite the team and everyone works in the same analyses, so there is no mailing versions around and no question about which file is current.
Because the three analyses sit side by side, cross-checking them is part of the work rather than a reconciliation exercise at the end. Ownership transfers when someone rolls off, finished work archives out of the way, and an activity trail records who changed what.
project · Meridian CDD
EU industrial software
market sizing
8 players benchmarked
competitive
Target P&L + valuation
company
shared with 4 members · every change logged
05 · Hypotheses
Assumptions you can see, challenge, and reuse
Forecasts are built from quantitative hypotheses (per-period growth impacts like “pricing hardens two points a year from 2026”) that attach to driver cells and accumulate.
Dial each hypothesis per scenario to compare base, upside, and downside in the same model. When a partner asks “what's driving 2030?”, the answer is a named assumption, not a cell reference.
forecast drivers
Pricing hardens post-consolidation
+2.0pp / yr on avg. price · 2026 →
2030 market: €5.9B +8% vs. base
06 · Scenarios
Scenarios you can compare with confidence
Each analysis carries its own scenarios: base, upside, downside, or whatever the deal calls for. A scenario is a set of dial settings over that analysis's hypotheses, so switching one recomputes it instantly, with no forked copies of the model.
Because all three are modeled the same way in the same project, the cases line up. The downside you build in the market sizing is the one you hold the target's downside plan against, and the assumptions behind both are named and visible.
market sizing · scenarios
recomputed instantly · no forked model
07 · Sources & Traceability
Every number keeps its receipts
Values carry assertions: where the data came from and what judgement was applied on the way in. Computed lines trace back to the inputs that produced them.
Models are built programmatically: the AI structures and populates, but the math is computed, never generated. No hallucinated arithmetic, ever.
avg. price · DACH · 2024
€1,240
3 sources