8 of 53 Bids Preceded Medly AI's 8m Round

1 of 8 Suppliers, Chosen From 53 Bids

Eight suppliers won a place in the UK government's AI Tutoring Tools Pioneers Programme in June 2026, selected from 53 competing bids. Medly AI, a London-based AI exam-prep and tutoring startup, was one of the eight.

Each winner received a research contract worth GBP 300,000, excluding VAT, running through 31 March 2027. The programme is a procurement pilot, not an award ceremony or a grant competition: suppliers are paid to prove their AI tutoring tools work under real public-sector conditions, with government evaluators watching the results, before any wider rollout decision gets made.

That distinction is what makes the timing worth noting. Medly AI had already survived a formal, competitive, publicly documented evaluation process months before its next funding round closed - a test no press release can substitute for. Fifty-two other bidders did not make the cut, which is what gives the eight winners' selection any evidentiary weight at all.

Two Months Later, an 8 Million Dollar Seed Round

Medly AI announced an $8m seed round on August 19, 2026, led by Felix Capital. Existing backers Eka Ventures and Ada Ventures both returned for the round, joined by a group of named angel investors: InvestEngine chief executive Andrey Dobrynin, Creator Fund venture partner Al Giles, LearnLaunch general partner Jean Hammond, and Episode 1 general partner Hector Mason.

The round is the company's second. Medly AI first raised GBP 1.7m in a seed round led by Eka Ventures in February 2025. Growing from a sub-2-million-pound opening round to an $8m round with a new lead investor inside eighteen months is a trajectory that normally rests on user growth and revenue metrics alone; here it also rests on a public-sector track record that simply did not exist at the time of the first round.

A Tender Did the Diligence a VC Usually Has To Do Alone

Felix Capital did not have to take Medly AI's classroom claims on faith. Weeks before the round closed, a government evaluation panel had already tested the same claims against 52 competing bids and was paying Medly AI to keep proving them under an active contract that runs into 2027, long after any due-diligence memo would normally have been filed away.

That is a different kind of signal than a reference call or a growth chart. A competitive public tender forces a vendor's product claims through an evaluation process it does not control, run by a buyer with no financial stake in the outcome. For an AI-education vendor, winning 1 of 8 slots from 53 bids functions as third-party validation that a private investor can price into a round without reconstructing that scrutiny in-house.

This is likely to become a repeatable sequence rather than a one-off. An AI-education startup that wins a competitive public pilot gets a cheaper, faster path to its next funding round: raise a modest seed, win the government tender, then raise the larger round on the strength of it. Medly AI's path from a GBP 1.7m opening round, through a government contract, to an $8m round led by a new investor is a playbook other AI-tutoring vendors are positioned to copy.

What This Means for Buyers, Not Just Investors

Any school trust, public-sector buyer, or employer evaluating an AI-tutoring vendor should treat "did they win a competitive public pilot" as a more reliable filter than the size of a funding round on its own. A large raise can reflect investor conviction in a market opportunity; a won government tender reflects an independent buyer's judgment that the product already works under real conditions.

The underlying UK tutoring market gives that filter real weight. Almost a third of 11-to-16-year-olds in the UK receive some form of private tutoring, an advantage that tracks family income more closely than it tracks academic need. Medly AI's founders, Kavi Samra and Paul Jung, both trained in medicine at UCL and worked as NHS doctors before applying neuroscience-informed methods and AI to GCSE and A-Level tutoring - a background evaluators can check on a CV, but a public tender win is the harder, external proof that the method holds up outside a pitch deck.

For any organization sourcing an AI-tutoring tool, the practical takeaway is to ask a vendor directly whether it has been through a competitive public evaluation, and to weight that answer at least as heavily as its most recent funding headline.

Read next: SK hynix Bets $28.6bn on Memory Prices Staying High  |  QuantumLight Doubles Its Fund to $500 Million