JobDropper Consulting
Daniel Negrila-Mezei
Senior Credit Risk Modelling Consultant
Bucharest, Romania
•Hybrid
•Senior
Overview
Senior Credit Risk Modeler and Consultant with hands-on experience in Basel/FRS9, PD/LGD/EAD and rating model development, validation and deployment.
CompanyJobDropper Consulting
LocationBucharest, Romania
Work typeHybrid
LevelSenior
Compensationnot disclosed
DomainCredit Risk
Proof items
Available nowDetails
Professional profile
Senior credit risk modeler and consultant focused on IFRS9, rating models and risk parameter estimation. Combines quantitative methodology development, SAS programming and .NET/C# tool development to deliver end-to-end modelling, validation and production solutions for banks and financial institutions.
Core expertise
- IFRS9 PD, LGD and EAD modelling and validation (including forward-looking components and stage allocation)
- Rating model development (corporate, SME/PI) and transition/migration approaches
- Statistical methods: OLS/LTS, quantile regression, Bayesian approaches, Vasicek one-factor, Markov-chain PD estimation
- Robust estimation and outlier handling (Least Trimmed Squares, bootstrap)
- Production-ready implementation: SAS (Base, Macro, STAT, ETS, IML, EG, Miner) and .NET/C# UI applications
Experience evidence
- Owner, JobDropper Consulting S.R.L. (since Nov 2022): delivered risk parameters estiamtion/validation methodologies and SAS solutions for Credex, Unicredit Bank(rating & LGD validation), Unicredit Leasing and Finance, Raiffeisen (IFRS9, collection scorecard) etc., support for BDO, Marsh, EY Romania.
- PwC Romania (Oct 2020 – Nov 2022), Senior Manager FSRR: lead quant and SAS programmer delivering rating models, IFRS9 parameter modelling (LGD/EAD OLS and LTS, FL adjustment), and stage-2 allocation (quantile regression) for UniCredit Group and others.
- KPMG Romania (May 2015 – Sep 2020), Manager Governance Risk & Reporting: main quant and SAS developer with group-level methodology input for UniCredit entities and IFRS9/rating model work.
- ING Bank (2013–May 2015), Senior Credit Risk Modeling: developed models, coordinated data collection and trained staff in SAS and modelling.
- BCR-Erste Bank (2008–2013): SAS programmer / Senior Statistician for risk parameters, provisioning, IRB input and real estate modelling validated by the National Bank of Romania.
- Earlier roles include Raiffeisen Bank and data-science work for P2P lending (Lend Rise USA).
Tools and methods
- SAS suite: Base, Macro, SQL, STAT, ETS, OR, IML, EG, Graph, Miner
- Programming and apps: Visual Studio / .NET, C#, C++, JavaScript, HTML/CSS
- Statistical methods: OLS, LTS, quantile regression, Bayesian methods, Vasicek, Markov chains, bootstrapping, copula simulation etc.
Education and certifications
- University of Bucharest, Faculty of Mathematics (1994–1998), License: Theory and Construction of Real Numbers
Career preferences
- Preferred work setup: Hybrid
- Focus areas: IFRS9, credit risk modelling (PD/LGD/EAD), rating models, model validation and consultancy projects
- Open to consultancy and technical methodology roles supporting banks or financial institutions. 
Gallery
10/104:49
High
Final score is the technical test score, completed in 4:49.
Expert / senior technical level
Card visibility
Visibility health
How strong this card looks for feed, search and browsing discovery.
7.0/10
Medium
Calculated from relevance, content quality, media proof, role fit, market benchmark and momentum. It is a discovery signal only; the final professional score stays driven by the technical test.
Fit
8.8/10
Core market signals are explicit: role, company, level, work type
Quality
8.4/10
Structured content is solid with 7 key fields completed
Market benchmark
7.9/10
Market benchmarked against 1 similar cards, enough to estimate local market position
Professional evidence check
Evidence strength + technical screening
Checks whether the description, domain signals and attachments support the declared role. The professional test remains the main validation signal.
WEAK EVIDENCE
35/100
The attachment thoroughly describes advanced methods for time series analysis and forecasting, focusing on the decomposition of time series into deterministic and stochastic components, model estimation, validation, and prediction intervals. While not specifically about Credit Risk, the statistical methods are applicable skills. The document shows high technical and workflow detail with moderate professional complexity indicative of senior level, but limited direct domain alignment with credit risk modeling concepts.
Analysis source: OpenAI
Evaluated: 13 Sep 2026 21:33. Responsibility evidence: zero means not demonstrated in the evaluated material, not zero ability.
Domain alignment40/100
Technical depth90/100
Workflow85/100
Responsibility evidence70/100
Attachments80/100
Validation75/100
Domain alignment: The content is not fully aligned with Credit Risk but rather focuses on advanced time series analysis, modeling, and prediction techniques, a specialization more typical of Data/AI.
Model logic: Domain-aware evidence model for Finance / Credit Risk: Alignment 32% · Technical depth 22% · Workflow 16% · Complexity 12% · Attachments 6% · Validation 12%. EvidenceBase 68/100 × DomainConsistencyFactor 0.72 = Final professional evidence score 35/100.
Evidence read: The attachment is a technical document detailing comprehensive methodologies for time series analysis, modeling, parameter estimation, and forecasting using statistical and mathematical models including ARIMA and spectral analysis.
Declared vs detected: Credit Risk · Data/AI
Weights: Alignment 32% · Technical depth 22% · Workflow 16% · Complexity 12% · Attachments 6% · Validation 12%
Level rationale: The attachment demonstrates advanced technical depth and workflow understanding suitable for a senior practitioner due to detailed description of complex models, statistical inference, and application examples with validation.
Test result: 10/10 · 4:49 · Expert / senior technical level · from 1 attempt
Recognized evidence
- Detailed and advanced coverage of time series modeling techniques (ARIMA, spectral methods).
- Clear explanation of deterministic and stochastic component decomposition.
- Describes model parameter estimation, statistical tests, and validation.
- Applicable statistical and mathematical modeling relevant to quantitative analysis.
- Includes practical example illustrating model application and validation.
Issues / risks / missing proof
- Content is general time series methodology, not specifically tailored to Credit Risk domain models like PD/LGD/EAD.
- No direct demonstration of credit risk concepts or specialized banking regulation models.
- Validation shown is general statistical, lacks industry-specific benchmarks.
- Workflow is methodological, but evidence of responsibility for decision-making or governance in credit risk context is missing.
Attachments
- The PDF attachment proves knowledge of comprehensive time series modeling and prediction methodologies.
- It lacks direct credit risk modeling context but is relevant to quantitative risk modeling and validation procedures.
- It showcases professional workflow elements such as model selection, testing, and forecasting.
- It provides measurable validation examples through statistical tests and confidence interval calculations.
- The document supports technical depth and workflow evidence but reflects a broader Data/AI focus rather than Credit Risk specificity.
- ARIMA.pdf: PDF supplied to AI for document analysis; local text extraction unavailable.
- JOBDROPPER_PLATFORM_REVIEW.docx: Content extracted and included in assessment (26248 characters; extraction limits apply).
Professional checks
- Presence of advanced time series statistical modeling (ARIMA, spectral analysis).
- Demonstration of model parameter estimation with least squares and orthogonal polynomials.
- Use of statistical hypothesis testing for model validation.
- Workflow evidence in model building, decomposition, and forecasting.
- Presentation of predicted intervals and error metrics for forecasts.
- No direct credit risk metrics or banking regulatory model references detected.
Recommended next
Similar cards ranked by domain, role, location and format fit.