Lime x High Intent Labs
Proposal
Automating the tender and RFP process with AI agents.
For Alan Clarke, Head of Public Policy EMEA, Lime
Thanks for making the time, Alan. This is our proposal for taking the manual grind out of Lime's tender responses: a team of AI agents that drafts, checks and prepares every bid, with your people signing off. Quick tour: who we are, who we work with, the team, the problem, how we solve it, and the offer.
Lime x High Intent LabsAbout us
High Intent Labs.
We build working AI agent systems inside companies, and train the leaders who run them.
Build
Agent systems
We design and build teams of AI agents that take over real workflows: drafting, checking, reporting. In production, not in demos.
Train
AI academies
We run AI academies for marketing leaders, product leaders and founders, in the UK and in Germany.
Who we are
Operators, not consultants
Fifteen plus years running growth, product and engineering at Trainline, Checkatrade, Auto Trader and CarNext.
Two things we do. We build agent systems that take over real workflows inside companies. And we train leaders to work with AI themselves, through our academies. Everyone on the team has run these functions in real businesses, which is why we build for production, not for slideware.
Lime x High Intent Labs
Our clients
Who we've worked with.
A mix of scale-ups and established names: Bolt, EY, Angi, Trip.com, Egon Zehnder, Mollie, Checkatrade, Blue Light Card and more. Mobility, marketplaces, finance, professional services. The common thread: real AI systems shipped into real operations.
Lime x High Intent Labs
Who you will be working with
Our team.
Sebastien de Bandt
Founder
Ex-CMO at Checkatrade, Super Payments and CarNext, Growth Director at Trainline.
Adrien de Cocatrix
CTO
Ex-CTO of Tracklabs and Beeldi. Leads how we build agents from start to finish.
Ali Goldsmith
Partner · AI for Marketing
Ex-Marketing & Brand Director at Auto Trader and Checkatrade. Runs the AI Academy for Marketing Leaders.
Rajinder Ashan
Partner · AI for Product
Ex-Senior Product Director at Trainline and CarNext. Runs the AI Academy for Product Leaders.
Nicolas Meibohm
Partner · Germany
Ex-founder and operator across media, mobility, biotech and AI. Runs the German AI Academy.
Jacob Haynes
Technical AI Director
AI researcher. Architects and scales advanced AI systems into production.
This is the team you would actually work with, no bench handover. Seb leads the engagement, Adrien and Jacob build the agents, and the partners bring the domain lens. Between us: Trainline, Checkatrade, Auto Trader, CarNext. We have shipped into regulated, operations-heavy businesses before.
Lime x High Intent Labs
What we will be solving
The problem.
Tenders decide where Lime operates. Today they are won or lost on manual effort.
The stakes
Challengers are winning
Competitors are taking tenders Lime should win.
The volume
Dozens every quarter
Across 29 countries, each with its own rules, forms and deadlines.
The cost
Manual and inconsistent
Time-consuming work under deadline pressure, with quality varying bid by bid.
Three facts drive this whole proposal. Challengers are winning tenders Lime should win. The volume is brutal: dozens of tenders every quarter across 29 countries. And the process is manual, so quality depends on who is available that week. That is exactly the shape of problem agent teams are good at.
Lime x High Intent Labs
How we solve it (illustrative)
From tender to submission.
1 · Read
Shredder agent splits the tender into questions, word limits and deadlines. Compliance agent builds the compliance matrix: every mandatory rule, form and annex.
2 · Draft
Six specialist agents draft their sections at the same time: safety & operations, sustainability, data & integration, community & equity, commercial & fleet, local policy.
3 · Check
Evaluator scores every answer against the city's own rubric. Evidence checks each claim against source of truth. Consistency catches contradictions. Below target goes back to the writer.
4 · Sign off
The bid lead signs off. Localisation and formatting applied. Nothing reaches the city without a named human approving it.
Built on Lime's answer library
Every past submission, approved claim and policy position. Every agent reads from it and writes back to it, so it grows with every bid.
Four stages. Read: two agents shred the tender into questions and build the compliance matrix. Draft: six specialists write their sections in parallel. Check: evaluator, evidence and consistency agents score everything against the city's own rubric, and weak answers go back automatically. Sign off: a named human at Lime approves everything. Underneath it all sits Lime's answer library, which gets smarter with every bid.
Lime x High Intent Labs
What we propose
The proposal.
A team of AI agents that drafts tender responses together, built and proven on a real bid.
24 h
First full draft of every question after a tender lands
6 wks
Proof of concept, on a live tender
£25k
Fixed fee, plus VAT. No surprises
Sept
We can start in early September
Delivered by a joint Lime + High Intent Labs sub-team: your bid knowledge, our agent engineering.
The offer in four numbers. Within 24 hours of a tender landing, a first full draft of every question. Proof of concept in six weeks, on a live tender, not a sandbox. Fixed fee of 25,000 pounds plus VAT. And we can start in early September. We run it as a joint sub-team, because the agents are only as good as the bid knowledge we feed them.
Lime x High Intent Labs
Next step
Let's put the first tender through.
One decision gets this moving: pick the tender we prove it on. Say yes this week and the agent team is drafting your first bid in September. Call or write any time.
