Selected Work · Keryx Labs

Reference · live

SectorTier-1 global bank

EngagementEnterprise AI architecture

Year2024 — present

Corporate Banking Payments AI & Analytics Platform.

Low-latency streamed payments data feeding a cross-platform governance and lineage architecture. Policy-as-code, signed audit ledger and a conversational AI interface deployed across the lines of business — powering live payments reporting, customer query management and incident recovery on the same evidenced substrate.

streaming-payments cross-platform lineage policy-as-code conversational AI audit ledger

100%

decisions signed

cross‑platform

lineage coverage

low-latency

streamed payments

Anonymised reference architecture · production view

Reference · live

SectorGlobal life sciences

EngagementAI/ML architecture design & delivery

Year2025 — present

Supply Chain & Operations Contract AI.

Conversational intelligence over a contract estate spanning hundreds of thousands of supplier, licensing and operations agreements. Every clause parsed, classified and made queryable in plain English by procurement, legal and finance — pricing exposures, indemnity terms and renewal clauses surfaced in seconds, with citation-backed answers traceable to the source paragraph. GxP-aligned audit trail across every query, response and citation.

contract-intelligence citation-by-default GxP-aligned conversational AI supply-chain-ops

100%

clauses citation-traceable

plain-English

query interface

GxP

aligned audit trail

Anonymised contract intelligence layer

Reference · scaling

SectorGlobal asset manager

EngagementOperating model · target architecture

Year2025 — present

Modernisation of Legacy Analytics Estate.

Migration strategy and operating model redesign for a large-scale data transformation across portfolio analytics, risk and client reporting. Iceberg-native lakehouse, contract-versioned data products, federated execution under central governance.

iceberg data-contracts operating-model portfolio-risk

data products

lines of business

100%

contract-versioned

Anonymised target operating model

Reference · live

SectorTier-1 UK bank

EngagementDesign & build

Year2021 — present

Customer 360.

Design and build of an enterprise-wide Customer Data Platform exposing externalised APIs for real-time identity, customer profile and event management. Rationalised 14 different identity-generating systems into a single resolved customer view powering personalisation and self-service across digital channels — contributing more than £20m in operational savings through automation, self-service and increased product upsell and cross-sell.

customer-360 identity-resolution real-time APIs event-streaming personalisation

14 → 1

identity systems rationalised

£20m+

operational savings

real-time

profile & event APIs

Anonymised customer-data-platform reference architecture

Reference · live

SectorTier-1 UK bank

EngagementStrategy, design & delivery leadership

Year2019 — 2023

Bank-wide Data & Analytics transformation.

End-to-end ownership of a bank-wide Data & Analytics strategy and £600m cross-franchise transformation programme — modernising the data estate from on-premise to a cloud-native lakehouse, embedding executive data ownership, and standing up a 30,000-colleague Data Academy. Delivered data-driven experiences for 19 million customers across retail, commercial, markets and wealth, with an estimated £400m revenue contribution from analytics-led interventions and a 20% reduction in run-cost.

enterprise-strategy cloud-migration data-governance data-academy multi-franchise

£600m

programme delivered

19m

customers reached

20%

run-cost reduction

Anonymised cross-franchise transformation blueprint

Reference · live

SectorTier-1 UK bank

EngagementML platform · architecture & delivery

Year2021 — present

Enterprise ML for customer value & financial-crime detection.

Established AWS-centric machine-learning and analytics capabilities at enterprise scale — powering a Customer Lifetime Value engine that drives revenue-positive interventions and balance increases, alongside automated transaction-monitoring and fraud-detection models that reduced risk and increased operational efficiency by over 60%. Same governance, same audit ledger, applied to the model lifecycle from training through challenge to production scoring.

aws-native customer-lifetime-value fraud-detection transaction-monitoring model-governance

60%+

efficiency uplift in fraud / TM

enterprise

scale CLV scoring

end-to-end

model governance

Anonymised AWS-native ML platform reference

Reference architectures, in production.

Corporate Banking Payments AI & Analytics Platform.

Supply Chain & Operations Contract AI.

Modernisation of Legacy Analytics Estate.

Customer 360.

Bank-wide Data & Analytics transformation.

Enterprise ML for customer value & financial-crime detection.

Want the unredacted version?