We spent the last few years studying where AI creates real economic leverage, which verticals are ready to be rebuilt, and what it takes to turn domain expertise into a category-defining company.
This is the framework we use to decide what to build, who to build with, and where to invest.
Current / past affiliations of LPs
How we pick · where we enter
−1
Expert + problem
0
Company exists
1
Customers
gAI enters here
Most venture models begin when a company exists. Some of the most interesting opportunities exist before that.
At minus one there is an expert who knows the problem and has no company. That is where we start.
The distance between −1 and 0 is where most industry insight dies.
How we pick · our lens
The interesting companies sit where all three overlap.
How we build · the engine
This is not a philosophy. It is a schedule. A founding team staffed, a workflow mapped, a production build shipped and a design partner signed, in about six weeks. The stack is repeatable across verticals, so we never rebuild the engine. We re-apply it.
Day 0
Expert identified
Week 1
Customer discovery
Week 2
Workflow mapped
Week 3
Prototype
Week 4
Production build
Week 5
Customer validation
Week 6
Design partner signed
Weeks to a shipped MVP · gAI vs the usual routes
150
→
a handful
We talk to hundreds of operators. We build with very few. Most ideas get killed.
That is not a failure of the system. That is the system.
How we build · what it has produced
Wealth · Financial services
$100M
advisor book moved in two weeks. Zero NIGOs.
AI for the complex workflows of wealth management, purpose-built for advisor transitions. 15+ live RIAs, including a $1B AUM firm. SOC 2 Type II.
“Not just in building the product, but in shaping it with deep technical expertise and vertical AI insight.”
Vineet Mohan · CEO, FastTrackr AI
Insurance · InsurTech
7 min
The review it replaced took 3 hours.
Restoration teams identify items and generate defensible replacement-cost values from photos, cutting contents estimation time up to 90%. 2 live customers.
“The technical depth, execution speed and strategic support we needed to focus our energy on the market.”
Don Milley · CEO, ContentsIQ
Lending · Financial services
3
design partners
The intelligence layer for small-business lending: find the right businesses, close loans faster, manage portfolios better. Works inside existing systems and controls.
Visit website ↗
Enterprise productivity
3+
live customers
The control plane for AI agents. Drag-and-drop skills and workflows any LLM agent can run, with cost, evals, security and access managed in one place.
Visit website ↗How we pick · what we are scoping next
Some of these will become companies. Some won't. This is what the top of our funnel looks like before an operator is attached.
How we pick · the markets we map
Eight sectors, mapped in public before we look for an operator: the structural pain, the number we are underwriting, and what we would build if we find the right person.
Financial services
4 markets
Accounting
75%
of accountants retiring within a decade.
Looking for: Agentic close and audit above the ERP.
Read ↗Insurance
60%
of operating cost is still manual work.
Looking for: Underwriting and claims that read unstructured data.
Read ↗Lending
$10B
in annual fraud losses.
Looking for: Origination and servicing as one intelligent layer.
Read ↗Wealth
$124T
changing hands in the largest US wealth transfer.
Looking for: An AI stack that scales advisors, not headcount.
Read ↗Enterprise
3 markets
Productivity
$1T+
of productivity spend sitting between people and outcomes.
Looking for: Systems that execute, not just summarize.
Read ↗AI adoption
$200B+
of enterprise AI budget, mostly stalled in pilots.
Looking for: The control plane that turns pilots into production.
Read ↗Agent identity
$50B+
identity and security market built for humans.
Looking for: The trust layer for agents acting with authority.
Read ↗Commerce
1 market
Agentic commerce
$8T+
of digital commerce about to be agent-mediated.
Looking for: Discovery, trust and payment rails for agents.
Read ↗We publish the maps before the outcome is obvious.
The people behind the process
Builder · Operator · Investor
AI · Product · Engineering
Investor · Studio builder
The full playbook
We'll send you the full process for how we pick, build and invest, the portfolio memo and the Start Fund materials. Then, if it resonates, a conversation.
Talk to Amit / Vijay →For LPs, family offices, angels and accredited investors exploring early exposure to vertical AI. A working playbook, not an investment brochure.
The future of AI is vertical
