Data, analytics & AI for financial services

Data you can govern. Numbers you can defend. Decisions you can explain.

Elucid Consulting Group is a Prague-based boutique working with banks and financial institutions across Europe on data governance, regulatory data, analytics and AI. Founded in 2021 by a practitioner with more than twenty-five years of hands-on delivery inside banks and financial institutions, so the consultant you meet is the consultant who does the work.

See the practice → Talk to us
  • 25+years of hands-on delivery in banking and financial services by the firm's principal
  • 2021Elucid Consulting Group founded in Prague
  • 7 yrslongest single embedded engagement, inside a Tier 1 banking group's data warehouse unit
  • 5countries worked in directly, plus the wider CEE region — UK · NL · BE · AT · CZ
The practice

Six areas, one data value chain.

From strategy and governance through platforms, analytics and AI, each area is led by someone who has done the job inside a bank rather than described it from outside. Every engagement is tied to a regulatory or business outcome you can name.

Discuss a brief →
  1. 01

    Data Strategy & Architecture

    Capability assessment, the gaps that matter, and a target architecture that scales with the business without accumulating technical debt along the way.

    Data strategyTarget architectureData meshLakehouseCloud migration
  2. 02

    Data Governance & Quality

    Operating models, catalogues and quality controls designed around how the organisation actually works, so they are still in use a year after we leave.

    Data governanceCollibraUnity CatalogMDMLineageBCBS 239
  3. 03

    Analytics & Business Intelligence

    KPI frameworks, reporting layers and self-service analytics that put the right figures in front of decision-makers quickly, on platforms people log into.

    Power BILookerTableauSelf-service BIKPI design
  4. 04

    Data Engineering

    Reliable pipelines, clean transformations and tested data products on modern cloud platforms, built so your own team can own and extend them.

    dbtAirflowSparkDatabricksSnowflakeAzure / AWS
  5. 05

    AI & Machine Learning

    Predictive models and generative AI applied to concrete business problems, explainable by design and engineered for production rather than for a demo.

    Predictive modelsGenerative AILLMsGenie AIMLOpsPython
  6. 06

    Programme Leadership & Advisory

    Senior cover for data transformations: interim CDO, head of data or delivery lead, chairing the committees and steering the stakeholders a programme depends on.

    Interim leadershipDelivery managementVendor selectionChange management
Casework

Five programmes, told plainly.

Anonymised, with the figures clients reported at close or track today. Each one was led by Elucid's principal, including engagements that predate the firm's founding in 2021. References are available on request within engagement terms.

01
5ECB audit obligations closed
Pan-European fleet management and mobility group, post-merger with a Tier 1 European bank
BCBS 239CollibraA-IRBData governance

Closing ECB audit findings and building a BCBS 239-ready governance framework inside a post-merger group.

A pan-European fleet management group, freshly merged into a Tier 1 banking parent, faced open ECB audit findings across data governance and its A-IRB return-to-compliance programme. Data stewardship was informal, KDE cataloguing was done in spreadsheets, and there was no unified view of data taxonomy across the merged entities. We embedded within the Enterprise Data Analytics function for three years. We established and chaired the Finance & Risk Data Committee — ten data stewards across five departments — and built the governance operating model from the ground up. Over 250 critical data elements were migrated and expanded in Collibra, each with full lineage, ownership, and data classification. A harmonised logical data model covering 75+ entities was designed to close the A-IRB data gaps identified by the ECB, and the sourcing strategy for work-out costs, concessions and write-offs was defined end to end. Five ECB audit obligations were closed. The DPD dataset from the acquired entity was integrated into the group risk engine, aligning 2M+ customer accounts and improving group delinquency accuracy by 30%. Data quality errors fell by 25% and taxonomy alignment between finance and risk improved by 40%.

02
3 days → 3 hrsreconciliation cycle for derivatives, bonds and loans
Tier 1 Central European banking group, 100+ business units across CEE
Enterprise DWHBCBS 239SecuritisationTeradataCOREP / FINREP

Seven years building the data foundation of a Tier 1 Central European bank.

A Tier 1 banking group operating across 100+ subsidiaries in Central and Eastern Europe needed to replace an ageing data warehouse that could no longer support group-wide regulatory reporting or withstand regulatory scrutiny. We were embedded inside the EDWH delivery unit for seven years, covering Basel III credit risk, finance, accounting and controlling domains. The enterprise data model was designed across 550 data structures and 10,000 data elements, reconciled to within 2% of EAD totals — the threshold required to obtain national regulator approval for the Teradata platform migration. Automated securitisation data sourcing from group entities eliminated more than 30 manual workarounds and brought the process inside BCBS 239 and Securitisation Regulation requirements. UAT and reconciliation of instrument data — derivatives, bonds and loans — was automated in SQL, cutting the reporting cycle from three days to three hours. The credit risk analytics workload was migrated to AWS, reducing infrastructure cost by 25%.

03
60 hrsof analyst time saved every month on COREP submissions
Major European financial institution, Brussels
Credit riskCOREPMoody's Risk AuthoritySQLPower BI

Group-wide credit risk reporting and cutting COREP submission overhead by 60 hours a month.

A major European bank needed to consolidate its credit risk reporting across group entities and reduce the manual overhead in its quarterly COREP submissions. The existing process relied on inbox-based sign-offs, fragmented collateral systems, and manual corrections that consumed over 60 hours of analyst time each month. We led the data architecture and reporting strategy for the Moody's Risk Authority platform, mapping counterparty credit risk data to COREP C34 template requirements and acting as the single point of contact with the platform vendor for defect resolution. More than 20 SQL-based data quality checks were built and automated, eliminating the manual correction cycle. Collateral management systems were consolidated into a single group-wide solution, reducing system redundancies by 50% and cutting credit risk data processing time by 35%. Power BI reconciliation workflows replaced manual sign-off processes, reducing stakeholder approval time by 80%. ECB regulatory discrepancies fell by 20%.

04
40%fewer compliance issues after four risk silos became one
The largest financial services group in the Czech Republic and Slovak Republic
Basel Pillar 2Risk architectureProgramme managementData consolidation

Transforming integrated risk architecture for Basel Pillar 2 across a Central European banking group.

The largest financial services group in the Czech Republic faced a strategic imperative to transform its integrated risk data architecture for Basel Pillar 2 — a programme that spanned Czech and Slovak subsidiaries and reported directly to the CRO. Risk data was fragmented across four separate silos, reconciliation was manual and error-prone, and the architecture could not sustain the reporting volumes required under the new regulatory framework. We led the programme end to end from Prague, consolidating the four risk data silos into a single unified data mart used for integrated Pillar 2 reporting. The programme delivered a 40% decrease in compliance issues, improved risk assessment accuracy by 25%, and established a single source of truth for risk data across both subsidiaries — an architecture that remained in place and continued to serve the group for years after the engagement closed.

05
80%fewer postings processed into SAP GL via the posting bridge
Tier 1 pan-European banking group, Credit Cards Tribe, Vienna
BCBS 239DatabricksUnity CatalogGenie AICredit cards

BCBS 239 remediation and AI-accelerated reconciliation on a Databricks Operational Data Hub.

A Tier 1 pan-European banking group needed to close ECB self-assessment findings across three BCBS 239 principles — accuracy, completeness and timeliness — within the Credit Cards Tribe, where data flows through a card management system, an Operational Data Hub (ODH) built on Databricks, and an enterprise data warehouse. The existing architecture had no formal lineage, manual reconciliation workarounds between the card sub-ledger and SAP GL, and no governed source-to-target mappings across the four operational card domains: acquiring, issuing, posting and collections. We are embedded as senior business analyst and data governance lead within the tribe. Unity Catalog was implemented across the full Databricks estate to provide column-level metadata management, lineage capture and access governance. Databricks Genie AI was deployed to accelerate data quality analysis and drive sub-ledger-to-GL reconciliation, reducing a multi-hour manual process to a supervised AI-assisted check and reconciling the credit card sub-ledger to within 1% deviation. A posting bridge algorithm was designed to consolidate and aggregate card management system postings into SAP GL, cutting the volume of postings processed by 80%. Settlement reconciliation with Visa and Mastercard was fully automated from source-to-target mappings, eliminating all remaining manual workarounds and closing the ECB findings.

Client names are withheld under engagement terms. Engagements before 2021 were delivered by Elucid's principal before the firm was established. Figures are client-reported at programme close, or taken from live programme tracking where the engagement is ongoing.

Sectors

Where the work has taken us.

Banking and financial services is the whole practice, not a vertical within it. The list below is where the work has actually been done, and what it involved.

Universal banks
Group-wide credit risk reporting, capital calculation, EDWH delivery, ECB audit response and BCBS 239 remediation.
Leasing & mobility groups
Post-merger data governance, A-IRB compliance, finance and risk data taxonomy, Collibra implementation.
Investment & corporate banks
Counterparty credit risk, securitisation reporting, COREP / FINREP, regulatory data warehouses.
Retail & consumer banking
Credit card sub-ledger reconciliation, DPD data, Basel Pillar 2, integrated risk data architecture.
Asset & wealth managers
Reference data, investor reporting, EMIR / SFTR transaction reporting and fund data platforms.
Insurers
Solvency II data assurance, IFRS 17 implementation, actuarial-to-finance reconciliation.
Specialist lenders & fintechs
Loan origination data, structured finance, securitisation reporting and scale-up data foundations.
Regulators & public sector
Data transparency initiatives, supervisory data platforms and compliance-driven analytics.
Delivered in United Kingdom Netherlands Belgium Austria Czech Republic Slovakia Wider CEE
How we work

Small on purpose. Senior by default.

We keep the firm small so that every engagement is led by a senior practitioner who stays for the duration, works inside your tooling, and leaves behind things your own team owns.

Evidence before instinct

We start from your data, your systems and your real constraints, not from a template deck. Every recommendation can be traced from source to outcome, and every assumption is written down.

Clarity at every layer

Architecture decisions are explained in business language and business requirements are turned into engineering specifications, without losing fidelity at either end.

Built to outlast us

The test of an engagement is what the team can do after it ends. We pair with your people, document as we go, and build platforms and processes that do not need us to keep running.

Where an engagement usually starts
  1. Discovery sprint2–4 weeks

    A senior-led assessment of a data strategy, a platform, AI readiness, or a programme that has lost momentum. Short enough to agree quickly. Specific enough to act on.

    A written diagnosis and a prioritised roadmap.

  2. Embedded deliveryMonths to years

    A data architect, analytics engineer, ML engineer or delivery lead inside your team, in your ceremonies, building in your systems. Knowledge transfers as the work happens.

    Capability and documentation your team owns.

  3. Advisory retainerOngoing

    A day or two a month alongside your CDO, Head of Data or CTO for vendor choices, architecture reviews, programme health checks and AI strategy.

    Senior counsel on call when it matters.

Contact

Bring us the problem.

A strategy question, a platform decision, a stalled programme, a team that needs senior cover. Send a few lines and you will get a straight answer on whether we can help.

Tell us who you are and what is going on. Short is fine.

Regulators, deadlines and what has already been attempted all help us reply usefully.

Enquiries are confidential. Our privacy notice explains how we handle your details.