Skip to content
thrv Portfolio

We start new companies using thrvAI and our Jobs-to-be-Done method.

We find complex customer jobs in large markets, build agentic intelligence products with thrvAI that get those jobs done faster and more accurately, and fund each new company from seed to product-market fit and Series A. If you are an investor, contact us to learn more.

From market to Series A

thrvAI analyzes hundreds of markets. The ones that meet our criteria become new companies.

  • Markets100s analyzed
  • thrvAI builds v1Seed funded
  • Product-market fitIterated with agents
  • Series AFirst round

Customers are not buying a product. They are hiring a product to get a job done.

Clay Christensen Harvard Business School

Who we are

We have refined our method since 2013.
Now we use it to start companies.

A method built with leading companies.

We developed our Jobs-to-be-Done method and platform with Microsoft, Google, American Express, eBay, Oracle, Twitter and Target. The method identifies the unmet needs in the customer’s job and the weaknesses of competitors, based on how fast and accurately their products help the customer get the job done.

A track record of creating equity value.

We have used our method to create equity value with the companies we work with. You can read the case studies and hear directly from the CEOs we have worked with.

Experienced operators and investors.

Jay Haynes, our founder and CEO, was an investment professional at GTCR, a product manager on Windows at Microsoft, and CEO of two startups, one funded by Kleiner Perkins and the other by DFJ. Jay leads the thrv team of executives from leading technology companies, supported by experienced advisors. Learn more about us.

  1. v1SaaS era
    2013
  2. v2SaaS era
    2015
  3. v3SaaS era
    2020
  4. thrvAIAgentic Platform
    2023
  5. thrv PortfolioStarting new companies
    2026

Our method

Technology changes.
The customer’s job does not.
This is a competitive advantage for our portfolio companies.

Theodore Levitt of Harvard Business School famously said, “Customers don’t want a quarter-inch drill, they want a quarter-inch hole.” This is true in every market. Products change with each wave of technology, from the CPU in the 1970s to the internet in the 1990s to AI in the 2020s, but the customer’s job stays the same. This is why the customer’s job is the North Star and competitive advantage for our portfolio companies.

Consumer

Consumers don’t want iPods.

They want to create a mood with music.

  • Vinyl 1964
  • Optical 1984
  • Hard drive 2001
  • Cellular 2007
  • Streaming 2021

Business

Salespeople don’t want CRM.

They want to acquire new customers.

  • Phone 1876
  • Email 1972
  • On-premise 1993
  • SaaS 2002

Health

Surgeons don’t want stents.

They want to restore artery blood flow.

  • Amputation 1800
  • Balloon 1964
  • Stent 1986
  • Drug-eluting stent 2002

The customer’s job

Every job has traffic jams.
Our portfolio companies find the alternative routes.

Customer jobs are complex. A typical job has 15 to 20 steps to plan, execute, assess, revise and conclude the job. Each step has 10 to 15 customer needs, for 120 to 150 needs in total, and every job has metrics, root causes, constraints and emotions.

Take the job of getting to a destination on time. Like every job, it has the equivalent of traffic jams that keep the customer from getting the job done. The customer has to make an important and difficult decision about whether to take an alternate route to arrive on time.

With thrvAI, our portfolio companies build the equivalent of full self-driving agents: intelligent agents that get the job done for the customer as automatically as possible to create economic value for the customer.

  • 20Job steps
  • 140Needs
  • 25Metrics
  • 35Root causes
  • 40Constraints
Example job5 of 20 steps

Get to a destination on time

  • Determine the routePlan
  • Travel to the destinationExecute
  • Assess if the arrival time is accurateAssess
  • Revise the route if neededRevise
  • Arrive at the destinationConclude

From job to strategy

The job is the market.
thrvAI finds the strategy for our portfolio companies.

We define the market by the customer’s job

We define each market for each of our portfolio companies by the customer’s job, not by the products. Customers want holes, not drills. The customer’s job is the market.

We size the market by willingness to pay

We pioneered and patented sizing market opportunities by the customer’s willingness to pay to get the job done rather than by the price of current products. This is how we size the markets for our portfolio companies. This same analysis is how we identify the target customer segment and the ideal customer profile for our portfolio companies.

We find the unmet needs and competitor weaknesses

thrvAI identifies the unmet needs in each step of the job and where competitors’ products fall short in getting the job done fast and accurately.

We set the agentic product strategy

thrvAI uses this analysis to set the optimal strategy for an agentic intelligence product that gets the job done and creates economic value for customers.

Investment criteria

What we look for when we start a new portfolio company.

With thrvAI, we are analyzing hundreds of business, consumer and health markets against five investment criteria.

01Complex jobs

Our first criterion is a large market with a complex customer job, where customers struggle to get the job done quickly and accurately with the competitive products in the market today.

02Difficult decisions

We look for jobs where it is difficult for customers to make the important decisions that determine whether they succeed.

03Intelligence + compute

We look for jobs where intelligence and compute can get the job done faster and more accurately.

04Large markets

We target large markets, sized by the customer’s willingness to pay to get the job done, the frequency of the job execution, the number of customers in the market, the economic value of the job, and the opportunity to disrupt the incumbents.

05Human impact

We look for jobs where getting the job done better has a meaningful impact on people’s lives. This is the criterion we weigh most.

The process

How we start a new portfolio company.

Step 01

We identify markets that meet our criteria

thrvAI maps the customer’s job-to-be-done, finds the unmet needs and competitor weaknesses, and sizes the market by willingness to pay. We are running this analysis on hundreds of markets.

Step 02

We start the company and build v1

When a market meets our criteria, we start the company, provide the seed funding, and build v1 of the new agentic intelligence product with thrvAI.

Step 03

We iterate to product-market fit

We put the product in the hands of customers and iterate rapidly with thrvAI agents until the company reaches product-market fit.

Step 04

We raise Series A capital to grow

Once our portfolio company has product-market fit, we raise Series A capital to grow the company with thrvAI.

What we build for each company

Full self-driving agents for the customer’s job

01A JTBD agent harness

The harness drives the whole job to completion. It maintains the state of the entire job, ingests updates, runs agents concurrently, and determines the next action. thrvAI creates the intelligence that guides the agents to refine their work and adjust the plan as they learn, carrying the job from request to result without the customer managing every step.

02An autonomous agent team

thrvAI maps the customer’s job and builds specialized agents to Plan, Execute, Assess, Revise and Conclude the job. Each agent owns a distinct part of the job, with the context, tools and expertise to get it done on the customer’s behalf.

03Evals that run in the background

thrvAI runs evals autonomously, continuously improving the economic value created for customers as measured by the metrics in the job.

04Error analysis on every failed run

When a run fails, thrvAI reads the trace, labels the failure, finds the first span where it went wrong, and fixes the error.

05An MCP tool surface

Each product gets an MCP tool surface, so its job agents can read and act on any of the customer’s systems of record.

06Open-weight models per job step

Proprietary open-weight models are trained for each job step on the company’s own job executions and expert preferences.

07Outcome pricing

Customers pay for the outcome of getting the job done, and the cost of each job execution is managed as a gross margin line.

“The application layer on top, ultimately, is where economic benefit will happen.”

Jensen Huang

Why we do it

Intelligent agents will get the job done for customers to create economic value and improve their lives.

This is the reason behind each of our investment criteria. The most important is human impact: helping customers get their jobs done better so they can live better lives.

Learn more about the thrv Portfolio.

If you are an investor, an executive, or an expert in a complex customer job, get in touch. We would like to meet you.