Skip to content
KS

SUMMARY

Process & Manufacturing Engineer (Aerospace) · Applied AI/ML — Modelling, Verification & Validation

Aerospace process engineer at Airbus Defence and Space with an independent applied-AI/ML research track centred on verification and validation. Three years developing technological processes, assembly sequences and control documentation for prototype and serial aircraft production; in parallel, building probabilistic models and reproducible validation pipelines — deflated performance metrics, combinatorial purged cross-validation, calibration and uncertainty quantification — and publishing the results (a self-archived preprint; an open-source research artifact with a Zenodo DOI). FEM-certified; CATIA V5, SAP ERP and Python in daily use.

EXPERIENCE

Sep 2023 — Present

PROCESS ENGINEER

Airbus Defence and Space

Warsaw, Poland

  • —Progression: Intern (Sep 2023) → Junior Process Engineer (Jan 2024) → Process Engineer (May 2026)
  • —Develop technological processes, assembly sequences and control cards for parts and subassemblies in prototype and serial aircraft production (Airbus Poland and CASA/Airbus cooperative programmes)
  • —Optimised production workflows and material/fastener allocation — up to 30% efficiency improvement and ~€50K annual cost savings
  • —Verify processes against new technical developments, inspection-monitor workstation compliance, and implement design-office changes
  • —Define tooling, fixture and metrology assumptions; select tools on techno-economic criteria; normalise working time
  • —Deliver internal training on design & technological documentation; apply Lean standardisation and FOD discipline daily (CATIA V5, SAP ERP, GD&T)
2024 — Present

CO-FOUNDER & CTO

K&S Venture Group — Clashpoint.me

Warsaw, Poland

  • —Designed, built and operate an AI-driven sales-coaching platform end to end — FastAPI, PostgreSQL, Next.js/React, LLM features

RESEARCH & PUBLICATIONS

Self-archived preprint · 2026

Deflated Sharpe ratio, a sequential adaptation of PBO (probability of forecast overfitting) and walk-forward retrodiction applied to published AGI capability forecasts — a pre-registered pipeline that reports its own failed prediction.

Open source · Zenodo DOI 10.5281/zenodo.20645678 (v0.2.0) · GitHub (github.com/Ricko12vPL/worldclass-research-org)

Multi-agent pipeline enforcing pre-registration and reproducibility across the research lifecycle; centrepiece pre-registered gate-calibration study, with a public erratum (v0.2.1) downgrading its cross-arm results to exploratory.

Solo research · 2024–2026

v1 (≈274K LOC) was retired in Aug 2026; its published metrics are withdrawn. v2 rebuilt the process under a written research constitution — pre-registration, an append-only decision register, a permanent attempts ledger and separate build/verify sessions. No strategy it tested was validated for trading, and none is recorded as reaching paper or live trading.

Private · in progress · 2026

Successor to X_Quant: point-in-time data, a fast vectorised backtest engine, a signal lab and a research-integrity engine (hash-chained trial ledger, sealed holdouts, DSR / PBO / SPA). Key phases close with an independent AI-agent review, and the integrity statistics are mutation-tested; no real-return results yet.

Manuscript in preparation

NumPyro hierarchical Bayesian classifier with a LightGBM baseline for near-Earth-object triage; calibration, uncertainty quantification and a reproducible training pipeline.

SKILLS

VERIFICATION & VALIDATION

Combinatorial Purged CVDeflated Metrics (DSR)PBO / CSCVWalk-ForwardCalibration & UQPre-registrationReproducible PipelinesSPC / Control Cards

AI / ML

Bayesian Modelling (NumPyro)LightGBM / Gradient Boostingscikit-learnPyTorchSHAPLLM & Multi-agent PipelinesRAG

MODELLING & SIMULATION

FEM — Linear & Nonlinear (certified)Convergence / Mesh-independenceCATIA V5Autodesk InventorPython numerical stack

MANUFACTURING ENGINEERING

Technological Process DesignAssembly SequencingTooling & MetrologyGD&TSAP ERPLean / Continuous ImprovementFOD

BACKEND & PROGRAMMING

PythonSQLTypeScript / JavaScriptFastAPIPostgreSQLNext.js / ReactDockerGit

EDUCATION

Oct 2024 — Aug 2026

Master’s programme — Project & Process Management

Collegium Civitas

Warsaw

Project-based

SOFTWARE ENGINEERING — 42 WARSAW

Peer-to-peer coding program

Warsaw

  • —Project-based, peer-to-peer software engineering — C, algorithms, Unix systems
Oct 2018 — Feb 2022

BEng Mechanical Engineering

Silesian University of Technology

Gliwice

  • —Thesis: feed mechanism of mining rigs and bolting machines · Major GPA 4.17 / 5.00

CERTIFICATIONS

FEA Convergence & Mesh IndependenceCoursera
FEM — Linear, Nonlinear Analysis & Post-ProcessingCoursera
Certified User: InventorAutodesk
Introduction to Orbital MechanicsUdemy
Interplanetary Spacecraft & Satellite EngineeringUdemy
Google Data AnalyticsGoogle
Python for EverybodyUniv. of Michigan

LANGUAGES & ELIGIBILITY

Polish (native) · English (professional working proficiency — daily working language at Airbus). Polish national (EU / ESA Member State). Holder of a Polish national personnel security clearance; experienced with export-control and classified-information regimes.