Exceptional Promise (Digital Technology) – honest review of my evidence list please

Stage 1: Tech Nation Endorsement
1

I’m preparing my Global Talent Stage 1 application under Exceptional Promise in Digital Technology and would really value honest, critical feedback before I submit. My areas are Semantic Engineering, Classical AI and Quantum Machine Learning. I graduated in May 2024.

Mandatory Criterion – recognised as having potential to be a leading talent

  1. NQCC Quantum Hackathon 2024 – Industrial Mentor. I was invited by the National Quantum Computing Centre to mentor a team of 8. I have testimonials from 2 team members describing how the mentorship influenced their decision to pursue quantum computing. Both are now doing PhDs in the field.
  2. Letter of Recognition – Oak Ridge National Laboratory (US DoE). For developing an Explainable AI framework for one of their solutions, a capability they didn’t have before.
  3. Compensation evidence. I’m a Semantic Engineering Team Lead at a fintech company, leading the semantic AI workstream on a project for a major financial institution. My salary is above the UK national average and above peers at a similar career stage.

OC1 – Innovation

  1. Patent application (Quantum ML). A quantum algorithm that prunes neural networks by ~50% while retaining ~99% accuracy. Developed under the UK government-funded QTAP programme hosted by Digital Catapult, which has a public blog post about the solution.
  2. Patent application (Systems Engineering). A requirements assistant that checks design requirements against organisational rules, plus other capabilities I can’t disclose publicly.
    Both were developed while I was employed at UKAEA, and I have an email from the UKAEA IP office confirming my role.
  3. Three UKRI-funded projects. I’m not named on the grant awards, but I have confirmation emails from the lead investigator describing my involvement.

OC4 – Academic contributions through research

  1. University of Oxford – Neuro-symbolic / Explainable AI project. Letter from a senior research professor detailing my contribution. The project included an event at Oxford attended by academics from MIT and Caltech.
  2. Imperial College London – Quantum AI for aerodynamic simulation equations. Journal paper submitted, preprint on arXiv. Letter from the Research Professor detailing my contribution.
  3. University of East London – Research Assistant, early Alzheimer’s detection. I used vision transformers on MRI scans with attention-map visualisation to localise affected regions, then integrated an LLM to generate reports and let clinicians ask questions about them.

Recommendation Letter: My three recommendation letter from the people who hold these titles

  1. Senior Quantum Technologist, Digital Catapult
  2. Vice President, SEGA
  3. Ex IBM executive

What I’d like feedback on:

  • Is my MC strong enough, and which pieces would you keep or drop?
  • Are OC1 and OC4 the right optional criteria for my profile, or would OC2 or OC3 be better?
  • Is OC1-3 (UKRI projects without my name on the award) worth including, or does it weaken the application?
  • Anything that looks like a red flag to an assessor?

Thanks in advance. Blunt feedback is very welcome.

2

Your MC has a plausible anchor in the Oak Ridge recognition, if the letter explains what you built and why it signals potential. The hackathon invitation helps if selective and based on your expertise; mentee testimonials show influence, but are weaker evidence of external recognition. Keep salary as context only. Senior recommenders add weight where each gives first-hand evidence of your work and standing; titles alone do not establish recognition.

I would retain OC1 and OC4 over OC2 or OC3 on this record. The two patent applications may support OC1, but unpublished status limits independent verification; the QTAP blog helps only if it names your contribution. The UKAEA email establishes involvement, not innovation or impact. OC4 is the more natural research criterion: use the Oxford work and the strongest documented Imperial or UEL contribution, with evidence of your own role and research output. The Imperial article is submitted, not published; the preprint is available, but describe it accurately. An academic letter should explain and endorse the research, alongside the underlying work.

Include UKRI projects only if confirmation emails and project records let an assessor verify what you personally did and the result. Being absent from award records is a real weakness, and unnamed awards do not prove your role; omit projects supported only by general investigator praise. OC2 needs evidence of its distinct requirement, which this account does not identify. OC3 is weaker still: a client workstream does not show significant impact at a product-led company. Prioritise the evidenced research and innovation, not all available projects.

3

Overall: you have real substance (UKAEA patents, Digital Catapult, Oxford, Imperial, ORNL). The risk is in how the evidence is packaged: too much of it is self-referential, unverifiable or not independent. Please check everything below against the current endorsing-body guidance, since the process has changed in recent years.

1. Is the MC strong enough?

The MC needs independent, third-party recognition. Ranked:

  • ORNL letter: keep, and make it your anchor. It is independent and from a prestigious institution. It must be signed by a named senior person and say what you built, that it was new to them, and the impact. If it only says “thanks”, it will underperform.
  • NQCC invitation: keep, but reframe. The evidence is that a national body selected you as an industrial mentor, so lead with the invitation and selection, not the outcome. The two testimonials from mentees are weak: they are junior, not independent, and the PhD point is about their decisions, not your recognition. Include them only as supporting material.
  • Compensation: this is your weakest item, and I’d drop it unless you can strengthen it. “Above the UK national average” proves nothing, because tech salaries sit well above the national average anyway. It needs a credible comparison to peers in digital technology at your stage, ideally with an employer letter. A better replacement is evidence that you were selected for the competitive, government-funded QTAP programme, or any award, press or invited talk.

2. OC1 + OC4 vs OC2 or OC3

  • OC1 is your strongest. Keep it.
  • OC4 is where I’d worry most. Letters and a preprint are not what assessors usually look for here. They want peer-reviewed publications, citations, reviewing roles or accepted conference papers. Do you have any accepted or published peer-reviewed papers? If so, put them front and centre. A submitted journal paper and an arXiv preprint are unrefereed and won’t carry much weight. The Oxford event with MIT and Caltech attendees is name-dropping unless a letter ties your specific contribution to it. For UEL, the question is what the output was (paper, deployed tool, citation?), not just what you did.
  • OC2 (recognition outside your day job) fits your mentoring, and possibly talks or community work. Consider it if you have more than the hackathon, but don’t use the same evidence in both the MC and OC2.
  • OC3 (a leading contribution at a product company) could fit your team lead role, but only with metrics (adoption, revenue, scale). Client work for a financial institution is often confidential and looks like consulting, so it is risky.

I’d keep OC1 and decide on the second criterion by which one you can support with the most independent, verifiable evidence. If OC4 can’t be backed by publications, OC2 or OC3 may be safer.

3. The UKRI projects (OC1-3)

I’d drop them from OC1. Not being named on the award is the problem, because assessors check public records and your name won’t be there. Emails from the lead investigator are self-serving and weak, and “involvement” in a project is not innovation. If you want to keep them, convert them into formal letters on university letterhead that name your specific role and link to the public Gateway to Research entry, and use them as supporting material for OC4 rather than as standalone evidence. As they stand, they dilute the strong items.