Ideas are easy. Choosing the first one is harder.
A leadership team can have dozens of AI proposals and still lack a clear starting point. Each department sees its own work, but the costs and dependencies cross the company.
Plek maps that work into a shared operating model. It connects roles to tasks, estimates the people and costs involved, then ranks where AI could make a difference. [1]
The company’s framing is direct: know where to put AI first. Named, priced and ranked. [1]
Start with the organisation chart
Plek starts with job titles, the number of people in each role and an organisation chart. Job descriptions can add detail; employee names and individual employee rows are not needed. [1]
Its model maps those roles to tasks and builds a first view of full-time equivalents, cost and AI exposure. The team reviews that first map and returns it within a day. [1]
That gives the leadership team a company-wide starting point without connecting an HR system or recording employee screens. [1]
The first map is built from roles and task research. Plek then checks the priority areas with the people who do the work. [1]
Ask the process owners what happens
Plek uses short AI interviews, usually scoped to one task or one job. The team can run them with the relevant owners in five to ten minutes. [1]
Those answers sharpen the map: what happens in practice, which steps repeat and where a proposed change would affect another team. [1]
Plek ranks opportunities by the cost of the work and how much of it a machine could carry. It can then help build and deploy the selected workflows with the company’s team. [1]
A connected knowledge graph lets people ask questions about the operating model. Teams can also query it through MCP, keeping the analysis tied to their own workforce. [1]
One company, or a whole portfolio
Plek works with enterprise teams, consultants and private-equity operators. The same model can show differences within a company or compare several companies using public workforce benchmarks. [1]
In one anonymised enterprise example, Plek mapped work across a distributed US infrastructure workforce of more than 13,500 people. The team identified priority AI opportunities in days. [1]
For a consulting engagement, Plek found a consumer-goods company’s selling, general and administrative organisation was 2.2 times the size of its closest competitor’s. That comparison helped identify where to investigate the operating model. [1]
Another engagement compared operating models across 15 companies in a $50 billion portfolio, giving the operators a shared way to prioritise transformation work. [1]
From work research to deployed changes
Founder Pierre Bouquet launched and led the MIT Work Analytics Lab. He previously built Biot AI in logistics and healthcare before its sale in 2022. [2]
Plek brings that research into the decisions companies make about where to introduce AI. The company joined Y Combinator’s Fall 2026 batch. [2]
The starting point is a map of the whole company, followed by conversations with process owners and specific work to deploy. Teams can book a demo to begin with their organisation. [1]
Plek’s goal is to carry the decision through: understand the work, choose the opportunity and make the change happen. [1]
Where Plek ranks
Plek isn't in a numbered ranking yet. Products get a position once enough companies compete on the same task.
Current rankings. Positions come from public evidence, and no company can pay for a place. How we rank
The bigger picture
The first AI project should have a clear job, an owner and a reason to come first. [1]
Plek is building the company-wide map that helps leadership teams make that choice and move it into practice. [1]
Written by SOTA2 from Plek’s public materials. Customer experiences and product plans are reported by Plek.