Application Review Request: Exceptional Promise (Senior Backend Engineer / AI Founder)

Stage 1
1

Hi everyone :wave:

I’m preparing my Global Talent Visa application under the Exceptional Promise route (Digital Technology, Technical pathway) and would appreciate feedback on my evidence structure before I submit in late June 2026.

Background

  • Current role: Senior Backend Engineer at a VC-backed UK automotive marketplace (Series C, Index Ventures, ~£1B+ in transactions processed)

  • Side project / Founder: AI platform company. Flagship product is an AI-powered safeguarding platform for domestic violence organisations, live in the UK and Nigeria

  • Experience: 5+ years across fintech, logistics, and AI

  • Technical focus: Node.js/TypeScript, event-driven architectures, multi-agent AI systems, AWS

Recommendation Letters (3)

  1. VP of Engineering, Current Employer — Direct line manager. Confirms senior role, scope of work (event-driven workflow engine, dispatch services at national scale), measurable impact on platform reliability, standing relative to peer group.

  2. NGO Partner Director (Nigeria) — Confirms adoption of my safeguarding platform by their organisation, real-world impact on GBV case management, names me as sole technical founder and builder.

  3. Senior Tech Industry Figure (TBC) — Confirming my contributions to the digital technology sector through product innovation and community engagement.

Mandatory Criteria — Recognition as a Potential Leading Talent (2 evidence docs)

MC1 — Employer Impact Statement

  • Letter from VP Engineering on company letterhead

  • Confirms Senior Backend Engineer role at a recognised UK tech scale-up

  • Specific contributions: designed and delivered event-driven workflow engine (XState + EventBridge + Lambda/SQS), dispatch services processing at national scale

  • Measurable impact: reliability improvements, latency reduction (specific metrics in letter)

  • Salary evidence as supplementary signal of market recognition

MC2 — AI Safeguarding Platform: External Product Recognition

  • Live AI safeguarding platform deployed across UK and Nigeria

  • 11 specialist AI agents (risk triage using DASH framework, safety planning, legal rights, service matching, etc.)

  • Multi-channel: WhatsApp, SMS, USSD, webchat — 16 languages including Pidgin, Yoruba, Hausa, Igbo

  • Innovative safety features: duress PIN with decoy wellness app, remote wipe, content warnings, caseworker wellbeing system

  • NGO case management portal with AI-generated risk assessments, case plans, outcome tracking

  • Evidence of adoption: NGO testimonial letter confirming live use + screenshots of production platform

  • Government ministry engagement for AI education initiatives

Optional Criteria 1 (OC1) — Innovation as Founder (3 evidence docs)

OC1.1 — Multi-Agent AI Platform (Technical Innovation)

  • Multi-agent orchestration platform with visual agent builder, 16 workflow node types, RAG pipeline, MCP support

  • LLM router with failover (Claude, GPT-4o, Amazon Nova), eval pipeline with LLM-judged gating

  • OpenTelemetry observability, RBAC, multi-tenant architecture

  • Architecture diagram + dashboard screenshots

  • Proves: founder of product-led digital technology company with shipped, differentiated technical features

OC1.2 — Safeguarding Platform Product Innovation

  • Problem: DV/GBV services overwhelmed, survivors can’t access help safely

  • Solution: 11-agent AI pipeline over 4 channels, 2 countries, 16 languages

  • Innovation specifics: duress PIN (no competitor has this), DASH-aligned AI triage, evidence vault with SHA-256 chain-of-custody, offline-capable portal with background sync

  • Comparison table showing no direct equivalent exists

  • Live URLs available

OC1.3 — Product Portfolio Breadth

  • Multiple shipped products across verticals:

  • AI safeguarding platform (UK + Nigeria)

  • AI education platform

  • AI-powered visa application builder

  • ML credit scoring for informal-economy traders

  • Shared infrastructure: proprietary agent SDK, observability command centre, multi-agent runtime

  • Proves: portfolio-level innovation, not a single product

Optional Criteria 3 (OC3) — Significant Technical Contributions (3 evidence docs)

OC3.1 — Current Employer: Event-Driven Workflow Engine

  • Technical write-up of the XState workflow engine designed and delivered

  • Architecture: XState state machines, EventBridge for events, Lambda + SQS for async workers

  • Before/after metrics on reliability and latency

  • Architecture diagram

  • Proves: significant technical contribution to a recognised UK tech scale-up

OC3.2 — Prior Product-Led Companies

  • Company A (fintech): SME virtual banking APIs, OAuth2/RBAC, Go microservices

  • Company B (procurement SaaS): 50% API performance improvement, 50% downtime reduction via Redis caching and observability

  • Named roles, named systems, measurable business outcomes

  • Proves: track record of significant contributions across multiple product-led companies

OC3.3 — Proprietary Observability Stack + Agent SDK

  • Built to operate a multi-product AI portfolio

  • Agent SDK: zero-dep, wraps Bedrock/OpenAI with auto-telemetry, health/error/LLM tracking

  • Command Centre: receives telemetry from all products, diagnoses and resolves incidents

  • Dashboard screenshots

  • Proves: entrepreneurial contribution — built horizontal infrastructure for the AI sector

Strengthening Evidence (2 docs, filling the 10)

Doc 9 — Community Engagement (July 2025 + May 2026)

  • July 2025: Invited by a Nigerian youth foundation to lead an AI bootcamp — 74 students, 4 schools, 3 government representatives (Federal Ministry of Labour rep, State Ministry rep, local education union chair)

  • May 2026: Invited to keynote and lead technical facilitation at a state-level AI hackathon — 100-150 students, partnership with state Ministry of Education + Ministry of Labour + the foundation, ₦300,000 prize pool

  • Launched an AI education platform to scale impact beyond one-day events

  • Proves: sustained community contribution over 12 months, with escalating scale and institutional recognition

Doc 10 — External Recognition

  • Tech4Good Awards 2026 nomination (AI for Good category) — backed by IBM, BT, HSBC

  • Technical blog post on multi-agent AI architecture (to be published on editorially reviewed platform)

Questions for the Community

  1. MC strength: Is the combination of VP letter from a recognised UK scale-up + AI platform with NGO adoption letter sufficient for mandatory criteria, or do I need stronger external recognition (press coverage, awards)?

  2. OC1 vs OC2: I’m currently going with OC1 (innovation) + OC3 (significant contributions). Should I swap OC1 for OC2 (community engagement) given my bootcamp/hackathon + education platform evidence, or is OC1 stronger with the multi-product portfolio?

  3. Founder evidence: My AI platform is strong technically but it’s my own company’s product. How do assessors typically view founder evidence vs employment evidence? Is an NGO adoption letter enough to validate it externally?

  4. Timing: The hackathon is May/June 2026 and I plan to submit late June. The bootcamp was July 2025. Is this enough historical depth, or does it look rushed?

Thank you for reading. I appreciate any feedback. :pray:

CC: @Raphael @alexnk @Francisca_Chiedu @pahuja

2

Hi @deji

Please avoid listing scope of work and role/responsibility confirmations in letters it comes across as you just doing part of your work and nothing extraordinary.

External recognition is considered by a national or international body. What your example confirms is your product adoption (which is not same as recognition). Engagement does not mean adoption and impact. The NGO letter needs to cover not just adoption but the actual impact of it. Industry impact and recognition needs to be baked in beyond adoption by multiple channels which just shows product scale. Overall MC needs to be stronger.

In all OC1, you need to establish what’s innovative about this, your role in the innovation, proof of product in market and associated market traction quantified (not self claims but third party evidence of this) along with validating support. Only self claims written and with diagrams have no proof of innovation and its impact beyond your own words.

Similarly in OC3, all documents seem to be purely self-claims. You need to have third party validation support of your claims. Secondly OC3 needs to show quantified impact in core company metrics of your contributions. Please get a variety of letters from customers who have benefitted and can validate the impact.

Overall only building, launching products and its adoption doesn’t qualify for industry recognition, innovation or impact. You need to show how it’s actually creating impact and especially on commercials. All validated beyond self claims. Only invitations and nominations don’t hold any value. Need to show participation and awards won. Evidences close to application date and especially one off are usually flagged. You need to show a track record of such activities

1 Like
3

@deji Your background as a Backend Engineer qualifies you for the Tech Nation application through the Technical Skill route. However, you have more than 5 years of experience. “Experience: 5+ years across fintech, logistics, and AI.” You will need to clearly explain the reason you are applying for promise in your personal statement and subtly reinforce it within some of your evidence.

On LORs

The VP of Engineering and the Senior Tech Industry Figure appear fine, but a letter from an immediate colleague or manager is also not sufficient. For the NGO Partner Director, he does not appear to be an industry expert. More so, Having referees cover different areas is good, but that is only one part of the requirement. The letters still need to follow the guidance’s LOR requirements.

Your MC:

It is not strategic to begin the Mandatory Criterion evidence list with a reference letter that is insufficient. It also appears that the letter is from the same line manager who wrote one of your LORs. The other items listed under MC read more like work descriptions than actual evidence.

Strengthening Evidence (2 docs, filling the 10): These are evidence kind of listings, but you need to specify which criterion you are using them for.

To be honest, your MC is very weak. OC1 are descriptions too.

I suggest reading the Tech Nation guidance carefully to understand what each criterion requires. Review how evidence is arranged on the forum, and take some time to gather more convincing, externally validated evidence to improve your chances.

All the best.

1 Like
4

Being assessed on the Exceptional Talent route means the MC bar shifts from “emerging national or international recognition” to “sustained national or international recognition.” That’s a meaningfully harder standard, and your current MC evidence doesn’t clear it on either version.

The deeper problem with MC1 is structural. Your VP of Engineering is already LoR #1. Using the same person as both a referee and your primary MC evidence document leaves assessors with one independent voice validating your industry standing. The MC needs external sources - people or organisations with no existing stake in vouching for you. Your government ministry engagement for the AI education initiative is a stronger signal than a line manager letter, but only if it comes as a formal acknowledgement from the ministry itself, not referenced in your own write-up.

Doc 10 is the one I’d fix before submission. The guide is explicit that nominations without wins carry no weight - “Tech4Good nomination” is not evidence of recognition, it’s evidence that you submitted your name for consideration. The blog post being “to be published” at application time is also risky; assessors flag evidence created close to submission. If the Tech4Good event has already taken place and you didn’t win or place, this slot needs replacing with press coverage where an editor independently decided your work was worth covering.

The OC1/OC3 split has a structural issue Pahuja’s point touches on. OC1.1 and OC3.3 are describing the same product from two angles - the multi-agent platform and its SDK. Assessors distinguish innovation from impact, but they also notice when two criteria are anchored to one piece of work. If you’re splitting OC1 and OC3 across the same platform, the separation needs to be clean: OC1 argues the architecture is technically novel relative to what exists; OC3 argues the deployment produced measurable commercial outcomes for an identified customer. Right now both documents read as architecture documentation. The OC3 evidence especially needs a customer or deployer letter validating what they actually experienced.

1 Like