01Primary
PE firms & portfolio companies
Pressure-test a target's AI claims before you close, then find and prove AI value-creation levers across the portfolio after.
Independent · Los Angeles
Strategy.Prototype.Proof.
(01) What this is
Executive product judgment on where AI pays off, and a working prototype that proves it before you commit the budget. For PE firms, portfolio companies and growth-stage SaaS.
(02) The gap
Most AI advice stops at the deck. Most AI shops start at the code.
Strategy firms hand you a roadmap and leave. Dev shops build whatever is in the ticket. Neither tells you which AI bet is worth making, or proves it before you commit the budget.
I do both halves: executive product judgment on what to build, and a working prototype with evals that shows whether it holds up.
- Strategy firmA deck× No proof it works
- AI dev shopCode× No view on what's worth building
- This practiceA decision + working proof
(03) Who it's for
Built for teams that need AI to pay off.
02
Mid-market companies adopting AI
You have document-heavy workflows and an AI mandate. I find the automations that pay back and prove them on your data.
03
Seed–Series B SaaS
Ship AI features customers will pay for, and price them so inference costs don't eat your margin.
(04) Engagements
Fixed scope. Published price.
- A · 2–3 weeks
AI Opportunity DiagnosticStart here
Workflow and product audit, a ranked AI Value Map, an ROI model, and one working prototype of the top opportunity. Fee credits toward next engagement.
from $7.5K
- B · 1–2 weeks
AI Product Diligence
Pressure-test a target's AI claims: data and eval maturity, token-level unit economics, roadmap risk, and value-creation levers.
from $15K
- C · 6–8 weeks
Prototype-to-Proof Sprint
Working POC with an eval harness, a business case and a build-vs-buy call. Handed off to your team or a vetted build partner.
from $35K
- D · 1–2 days / week
Fractional AI CPO
Embedded product leadership: roadmap, pricing and packaging, AI feature prototyping, team operating model, board narrative.
from $8K/month
- E · Weekly or biweekly
Advisory
A standing session for AI product and pricing decisions, plus async access between sessions.
from $2.5K/month
- F · Quarterly
PE Portfolio Retainer
A bank of diagnostic, diligence and advisory capacity to deploy across portfolio companies, plus a quarterly AI briefing.
from $15K/month
No hourly billing. Every engagement ends with something you can run, not just read.
(05) Process
Four phases, one decision.
Phase 011 / 4
Diagnose
Interviews, workflow and data audit. Rank AI opportunities on value against feasibility.
→ AI Value Map
Phase 022 / 4
Prototype
Build a working proof of concept on your real or sanitized data, with an eval harness from day one.
→ Working POC + evals
Phase 033 / 4
Decide
Accuracy, cost per task and a build-vs-buy recommendation your leadership can sign off on.
→ Business case
Phase 044 / 4
Hand off
You keep the code, the evals and the playbook. Production build goes to your team or a vetted partner.
→ Repo + playbook
You keep the prototype, the evals and the playbook.
(06) About
Eighteen years between the business and the engineers.
Josh Haas
I've spent 18 years building, pricing and scaling software products: e-discovery at Relativity, strategy consulting at EY, co-founding and selling Plannr, then product leadership at Welltok, Apto, Observer Media and Altvia.
Most recently I was Managing Director of AI Products at DLA Piper, where I took an AI platform from zero to $3M ARR in its first year, led a 50-person org across platform, research, evaluation and forward-deployed engineering, and cut private equity document reviews from 15 hours to 30 minutes.
Now I work independently with a small number of clients at a time, so the person who scopes the work is the person who builds the prototype.
- MBA in Finance, University of Denver
- PMP
- Pragmatic Marketing Certified
(07) Fit
Honest about fit.
Good fit
- ✓PE firms with an AI thesis, or portfolio companies under pressure to show AI value
- ✓Document- or workflow-heavy businesses: legal, financial services, real estate, insurance
- ✓SaaS companies adding AI features or repricing for AI
Not a fit
- ×You need a full production engineering team (I'll refer a partner)
- ×Pre–product-market-fit consumer apps
- ×You want someone on-site five days a week
Have an AI bet you can't size?
30 minutes. No pitch deck. If I'm not the right fit, I'll say so.