Riplo wants to sell consultancies the AI McKinsey built in-house

On 31 March, Riplo, a London-based company building software for consulting firms, came out of stealth with a €2.6m (£2.3m) pre-seed round, Startup Rise reported. Cherry Ventures, the Berlin investor, led it, with Blue Lion Capital, the founders of QuantumBlack and angels from OpenAI and Goldman Sachs among the other backers.

Riplo was founded in 2025 by Tobias Haefele, its co-founder and chief executive, with Oliver Scott and Zack Zornitta. Companies House shows Riplo Limited incorporated on 10 December 2025, with each of the three founders holding between 25% and 50% of the shares. Haefele and Scott met at QuantumBlack, McKinsey’s AI division, where they developed AI-native products, according to Startup Rise. They later reconnected at Hg Capital and met Zornitta there, an engineering leader with experience of scaling AI-driven software.

Riplo calls the product an agentic operating system for consulting. Its own site, as it stood on 31 March, describes a canvas that brings a project’s materials together, “creating a knowledge graph for every engagement”, from which consultants and agents produce client-ready deliverables. Startup Rise reports that the agents sit inside the consultants’ workflow rather than beside it, drafting, structuring and refining work as the consultants go. Tech.eu describes the interface as agent-first.

The largest consultancies have built their own versions of this. McKinsey launched Lilli, an in-house generative AI platform that searches and synthesizes the firm’s stores of knowledge, in August 2023. Riplo’s opening is the advisory and consulting firms below that tier, with the same proposal and slide workload but no in-house AI unit, which would have to buy such a system rather than build one. Two of its founders developed AI products inside McKinsey’s own AI division, and that record is much of what the round backs.

Backers from inside the big firms

Several backers come from McKinsey and its AI division. The founders of QuantumBlack, where Haefele and Scott met, are in the round, though neither report names them. Startup Rise lists angels from McKinsey, BCG, OpenAI, Apple, Goldman Sachs and Hg Capital, alongside senior figures from technology and professional services.

Haefele’s pitch is aimed at the format of the work itself rather than at the consultant. He put the consulting industry’s size at $1tn (£740bn) and likened its tools to a typewriter, Tech.eu reported. “You cannot achieve agentic speed inside a slide deck designed in the 90s,” he said in comments carried by Startup Rise.

Haefele said in the same statement that Riplo is not building another chatbot. Startup Rise reports that the company was set up because consulting tools are still built for manual, human-only work such as slide creation and spreadsheet modeling, and are hard for AI to operate within. Riplo’s answer is to put the agents inside the deliverable and keep what they produce: its site describes a shared “firm memory layer” to which every engagement adds.

What the money pays for

Riplo plans to expand its engineering team, deepen its work with design partners and keep building the platform, according to both Tech.eu and Startup Rise. Neither outlet names a design partner or a paying customer, and neither gives a revenue, customer count or usage figure. On 31 March, Riplo’s own site still invited firms to join a waitlist.

What can be checked is the amount, the backers and the founders’ record at QuantumBlack and Hg Capital. Tech.eu gives the amount as £2.3m and Startup Rise as €2.6m (£2.3m), the same sum in two currencies.

A market defined by who can’t build

For the managing partners of mid-sized consultancies and advisory boutiques, Riplo is a test of whether the gap between the largest firms and everyone else can be closed by buying rather than building. McKinsey could afford to build Lilli for its own consultants. A firm of 50 or 500 consultants is unlikely to have engineers to spare for that, and its alternative has been general chat tools that sit outside the deck and the model.

A system built as a firm’s shared memory becomes the place where its accumulated know-how lives. Riplo’s site pitches buyers on winning more work and retaining more clients, and answers the obvious worry about client material with single-tenant deployments, audit trails and technical controls “to prevent any data leakage”.

Riplo has the founders and the investors that a consulting buyer would want to see. What it hasn’t yet shown in public is a named consulting firm running its proposals and decks through Riplo. Until it names one, the case for Riplo rests on who built it and who backed it.