Help & Supportcareers@gravityer.com
AI is analyzing your overall score…
Identifying your key strengths…
Evaluating your skill match against the job requirements…
Assessing your cultural and operational fit
Aligning product with AI systems (prompts, evals, RAG, agentic) · Upheal · founder @ Lexomat · BNP Paribas
Generalist AI engineer with 9+ years shipping ML, NLP, and LLM products end-to-end. Most recently led prompt engineering, evals, RAG, and agentic systems for AI-generated clinical notes at Upheal — a Best Startup Award AI documentation platform for therapists. Before Upheal: physical climate-risk models at Cervest (Global Impact 50 Award), production financial NLP at BNP Paribas (a dozen NER models running daily in FX execution), MSc Machine Learning from UCL, BSc Mathematics from Imperial College London. Founded Lexomat (lexomat.sk), an AI legal-research chat for Slovak law and jurisdiction indexing millions of documents — drove product, design, agentic AI flows, GDPR, pricing, and infra end-to-end. I also co-founded a food delivery app in the pre-agentic-AI era, still used today. Obtained strong product, customer, and design instincts from owning both features and a company. Comfortable across prompts, evaluations, MLOps, design and full-stack Python / TypeScript. Currently taking on a small number of consulting engagements (LLM systems, eval frameworks, agentic, MLOps audits) and open to the right full-time or co-founding role at a serious AI venture. DM if either fits.
UCL
Master’s Degree, Machine Learning
January 1, 2016 – January 1, 2017
Imperial College London
BSc (Hons) & ARCS, Mathematics with Statistics for Finance
January 1, 2013 – January 1, 2016
Bilingválne gymnázium Milana Hodžu
High School, Mathematics and Information Technology
January 1, 2008 – January 1, 2013
Upheal
Senior AI Engineer
November 1, 2023 – Present
Prague, Czechia · Remote
ATEN Consult
AI & Product Consultant
August 1, 2023 – October 1, 2023
Greater London, England, United Kingdom · Remote
Cervest
Senior Data Scientist
July 1, 2021 – July 1, 2023
Greater London, England, United Kingdom · Remote
BNP Paribas
Machine Learning Researcher
June 1, 2017 – June 1, 2021
London Area, United Kingdom · On-site
Kiwi.com
Data Analyst/Machine Learning Intern
August 1, 2016 – September 1, 2016
Brno, South Moravia, Czechia · On-site
Imperial College London
Research Intern – MCMC on SDEs with unbiased estimators
June 1, 2015 – August 1, 2015
London Area, United Kingdom
Using Bayesian Networks to Make Inferences on Ebola Candidates
May 1, 2015 – June 1, 2015
Starting with a concise introduction to Bayesian Network and Decision Trees, the content primarily depicts a mathematical model to effectively decide whether an individual patient should be given Ebola treatment. In fact, two sub-models were interconnected to produce the optimized decision-making, namely a Bayesian network of associated nodes to Ebola and a Decision tree model involving all scenarios and corresponding costs. The written version was followed by a 20 minutes presentation, with an overall grade 86%.
Hénon map, Runge-Kutta Method applications
March 1, 2014 – Present
I analysed the behaviour of a discrete-time dynamical system called Hénon map, particularly with respect to parameter alternations. Next, I solved differential equations by numerical fourth-order Runge-Kutta method, pointing out its effeciency by comparing the results to analytical solutions, and applying the method to visualize the swinging Atwood’s machine motion, given by a system of differential equations.
Fractals and analemmata
February 1, 2014 – Present
Generated fractals using various conditions (e.g. equal probabilities of randomly chosen point of the fractal), created fully vectorized functions in Matlab to produce images of position of the Sun with respect to a fixed location on Earth/Mars (solar analemma for Earth/Mars) depending on planets' physical conditions (orbital eccentricity, the angle between spring equinox and perihelion, and the axial tilt).
Sestinas and Queneau Numbers
December 1, 2013 – Present
Sestinas (a type of Mediaeval Europian poems) are described by the rhyme scheme that can be represented via a sestina permutation function. The aim was to construct and test this function in Maple without any dynamic allocation of memory, verify its "complete cycle" property, and use its extended version to prove or disprove the validity of some natural numbers to be Queneau Numbers.
Euclid’s algorithm
November 1, 2013 – Present
I constructed Euclid’s algorithm to verify Fermat’s Little Theorem and Euler’s Totient Theorem. Using Maple, I also developed a safe communication system with another party by encoding and decoding messages via RSA encryption algorithm.
Curves on a Plane
November 1, 2013 – Present
Bézier curve is a parametric curve used to model smooth curves that can be scaled indefinitely, and are intrinsic to today's computer graphics (contour plotting, scaleable typefaces, CGI, etc.). The main task was to create a function for cubic Bézier curve and use for fitting the unit circle, firstly by approximating four quarter-circles and later six 60°arcs.
Monte Carlo Integration Methods
June 1, 2013 – June 1, 2013
The project provides a succinct insight into Monte Carlo methods, their history and main applications, followed by a comparison of stochastic integral approximations to numerical ones. The main objective is differentiation of two basic Monte Carlo Integral Approximation Methods: Hit-or-miss and Sample-mean methods, where the respective time-efficiency, convergence, algorithms and variance are particularly discussed. Followed by multiple presentations to mathematics-oriented audience, I earned 95% overall.
Databricks Certified Machine Learning Associate
Databricks
June 24, 2026 – Present
Design Patterns with Python
Lynda.com
June 24, 2026 – Present
Up and Running with Python
Lynda.com
June 24, 2026 – Present
Expert Certificate in Mathematics (Odborná štátna jazyková skúška z anglického jazyka; CEFR C1)
Jazyková škola Žilina
June 24, 2026 – Present
Python 3 Essential Training
Lynda.com
June 24, 2026 – Present
Next.js & React - The Complete Path
Udemy
June 24, 2026 – Present
Apache Spark™ Programming with Databricks
Databricks
June 24, 2026 – Present
Academy Accreditation - Databricks Lakehouse Fundamentals
Databricks
June 24, 2026 – Present
Expert Certificate in English (Základná štátna jazyková skúška z anglického jazyka; CEFR C1)
Jazyková škola Žilina
June 24, 2026 – Present
Cultural Fit Analysis
The candidate's project history, ranging from theoretical mathematics to practical AI applications, indicates a strong intellectual curiosity and a drive for continuous learning. Their experience in diverse environments (startups, large corporations, consulting) suggests adaptability. The focus on delivering impactful products and optimizing performance aligns well with a results-oriented culture. The personal projects, while not directly using modern ML frameworks, showcase a deep mathematical and analytical foundation, which is valuable for complex ML engineering challenges.
Soft Skills & Operational Fit
The candidate demonstrates strong leadership, problem-solving, and analytical skills through their project descriptions, especially in owning product quality and cost optimization. Their experience in mentoring new hires and reporting to founders indicates good communication and stakeholder management. The project diversity suggests adaptability and a proactive learning attitude.