The founder facet
My companies are the lab. Your P&L is the point.
I've run my own P&L with my own capital at risk. In a hiring context that's the proof, not the distraction: the systems below run unattended by design, they don't need me at a desk, and everything they taught me is what I bring to a seat.
Founder & CEO · 2025–
ClearForge.AI
AI deployment for mid-market and enterprise, strategy through production, one accountable team. Flagship: the go-to-market engine at a $4B industrial manufacturing company: 358 qualified targets, 119 account playbooks, 60% prospecting-time reduction.
clearforge.ai · deep dive below →
Co-founder & CEO · 2023–
Paradosi Partners
Family-office backed, with committed capital and three LOIs signed. A 157,000-company origination engine I built myself, investor reps, not just advisor reps.
paradosipartners.com · deep dive below →
ClearForge.AI · what I've actually shipped
Bain rigor, operator speed. The deliverable is the working system.
Most consulting value dies in the gap between recommendation and realization. ClearForge's product is that gap: I scope to the two or three levers that carry the value, then build, deploy, and leave running infrastructure behind.
The flagship · a $4B industrial manufacturing company
The go-to-market engine their sales team now runs on
This is the flow of strategy recruiters ask about: market scan to segmentation to qualified targets to account playbooks to a weekly delivery cadence, every stage automated, governed, and still running. Click through it.
Stage 3 of 5 · Qualified targets
A ranked, reasoned target list, not a purchased one
Each account is scored for fit and carries the reason to call: the segment it belongs to, the signal that makes it timely, and the recommended play. Prospecting time fell 60% because sellers start at the conversation, not the research.
Precision fastener distributor
Great Lakes · MRO consumables
Signal: New plant announced; no incumbent supply agreement visible
The play: Lead with a consolidated-supply program tied to the plant-opening timeline
HVAC components OEM
Southeast · OEM components
Signal: Reshoring supplier base; quality escapes cited in trade press
The play: Open on certified-material traceability and dual-sourcing risk
Rail maintenance contractor
Midwest · Infrastructure services
Signal: Multi-year federal program awarded; procurement cycle starting
The play: Map the award timeline and bid the first package, not the master deal
Illustrative sample of the system's output · client data blinded
AgentForge, the infrastructure underneath
internal · running four business units
The 35,000-line production multi-agent system behind all of it: 20 agents, a 9-tier model router, evaluation gates, and governance I answer to as the owner. The live agent on this site's hub runs on it.
Paradosi Partners · how I raised it and how I run it
Investor reps, not just advisor reps.
- The raise.
- Family-office backed, with committed capital behind a thesis my partner and I wrote ourselves: acquire one great founder- or family-held industrial business in a succession situation, move in, and operate it.
- The origination machine.
- I built the entire sourcing system myself: industry theses written against growth and succession dynamics, an embedding model that ranks the whole market for buy-and-build fit, outbound email on dedicated warmed domains, and a dialer that prioritizes owner-direct numbers with AI enrichment resolving owner names at roughly a 90% hit rate. Brokers and the personal network run alongside the direct engine. The funnel below is the system's live output.
The funnel · from market map to signed LOIs
5,200+
Industries evaluated
Growth, fragmentation, and succession dynamics scored per industry
18
Investment theses approved
A written buy-box per industry: why now, what good looks like, walk-away lines
157,000+
Companies identified and ranked
Every company scored by an embedding buy-fit model I built, before a human touches the list
73,000+
Passed screening
Size, ownership, and fit filters against the thesis
14,000+
Owners contacted directly
Warmed-domain email engine plus an owner-direct dialer with AI name enrichment at a ~90% hit rate
1,000+
Underwritten
IC-grade three-statement LBO models from a 17-tab engine I built
25
CIMs into full diligence
CIM deconstruction, validation gates, management meetings
3
LOIs signed
Direct-to-owner negotiations, no banker between us
Live counts from the sourcing system · bar widths stylized, numbers real · every stage of this engine is software I built
- The underwriting.
- 1,000+ targets screened and underwritten with IC-grade, formula-native three-statement LBO models produced by a 17-tab model engine I built: DCF, seller notes, real-estate treatment, sensitivity tables, the models an investment committee actually interrogates.
- The diligence.
- For the acquisition now under LOI, I ran buy-side diligence the way an institutional deal team would: full CIM deconstruction, document extraction into a structured deal-room summary, a three-statement LBO with explicit kill-switch criteria and lender-structure rules, three validation gates before any number reached the model, revenue-quality and margin checks against source documents, management meetings, and a 68-slide buy-side investment memorandum with open diligence gaps flagged rather than papered over.
- The deals.
- Direct-to-owner negotiations, no banker between us. Three LOIs signed to date, with the lead acquisition now in closing diligence. Targets and terms stay confidential until close.
Family-office backed · Committed capital · Owner-direct origination · IC-grade diligence · 3 LOIs signed
Why this matters to a hiring committee
Both ventures are staffed by AgentForge, the production multi-agent system I architected. That's the flight-risk answer in one sentence: the businesses run without me, and the operating muscle they built is exactly what the seat gets.
Would he fit your table?
About James →