About Josias
As a senior scientific R&D and software engineer, I bridge applied physics, mathematical modelling, and high-performance software to turn complex ideas into robust, production-grade systems. Over the past decade I have:
- Designed modular architectures for large-scale fare/ticketing engines and delivered mission-critical embedded software running across hundreds of field devices
- Built simulation and validation toolchains that accelerated integration and de-risked delivery
- Led quality transformations through advanced testing and CI, contributing to a 99% drop in production incidents
- Managed dozens of concurrent initiatives under Agile/Scrum
- Mentored cross-site teams to shared engineering standards
Domains & strengths: numerical optimisation and PDE-based modelling, 2D/3D & volumetric imaging, scientific computing, systems design, and performance tuning—applied across transport, healthcare imaging, and SaaS environments.
Communication & impact: former Master-level teaching assistant; author of extensive technical documentation; comfortable aligning research goals with operational constraints and executive priorities.
French
Native or bilingual
English
Native or bilingual
Remote only
Primarily works remotely
Experience
- JOSIAS ELISEE E.I.Scientific Software ConsultantRESEARCHMarch 2026 - Today (3 months)Valence, FranceI accelerate scientific and engineering software by combining numerical modeling/simulation expertise with rigorous software engineering and pragmatic AI/ML know-how. I typically operate in short, tightly scoped missions: clarifying the problem, building a PoC, validating outputs, and upgrading the stack (performance, architecture, CI/CD, reproducibility, documentation).Core capabilities include 2D/3D simulation, performance optimization, LLM fine-tuning strategy and evaluation, and scientific quality auditing (methodology, assumptions, validation criteria, and result consistency). I also support teams with technical documentation and hands-on training to make solutions maintainable.
- Career breakCareer transitionEDUCATION AND E-LEARNINGNovember 2024 - February 2026 (1 year and 3 months)Valence, FranceFull-time upskilling and hands-on prototyping across ML, LLMs, and agentic tooling targeting independent consulting- Designed and trained time-series prototype neural networks as part of an optimization re-training initiative (Python/NumPy/PyTorch on CUDA) with data-quality checks, imbalance handling, and validation workflows- Stood up local + hosted LLM stacks; experimented with quantization/decoding to balance latency, cost, and quality- Building a Python/LangChain application handling multilingual inputs beyond context size combining NLP and LLM technology
- ASTEKIngénieur Consultant SeniorTRANSPORTATIONJanuary 2021 - November 2024 (3 years and 10 months)Guilherand-Granges, FranceClient company: Conduent Business SolutionsEngaged in a multi-year mission to overhaul and innovate within the core ticketing and fare-engine domain for major public transport projects. My role evolved from pure architecture to hands-on development of mission-critical systems and technical leadership.- Architected and prototyped a scalable, multi-tenant C# fare-engine architecture designed for configurability and future growth- Delivered mission-critical embedded C++ software now in production across hundreds of validators on the Paris public transport network as part of a €1Bn+ infrastructure project- Spearheaded the creation of an innovative C++/CLI device simulator, which was championed by the client and adopted as the primary validation tool, significantly accelerating testing and integration phases- Mentored and provided technical leadership for developers across international teams (France, Australia), fostering best practices in complex project environments- Drove a culture of quality and knowledge sharing by authoring over 80% of the team's entire technical documentation from the ground up, creating a lasting asset for the organizationMy expertise and impact were recognized with an offer for a senior-level promotion, reflecting the value I delivered even amidst significant project challenges and pivots
Recommendations
Be the first to recommend Josias
Help this freelancer shine by sharing your experience working together.
These freelancer profiles also match your criteria
Agatha Frydrych
Backend Java Software Engineer
4.7
(3)
2
Baptiste Duhen
Fullstack developer
4.6
(4)
5
Amed Hamou
Senior Lead Developer
4
(2)
7
Audrey Champion
Web developer
4.3
(3)
4
Education
- PhD Computational PhysicsUniversity College London2011• Internationally recognized research in mathematical and physical modelling (published in peer-reviewed journals) • Acquired advanced numerical methods development skills in the: - Optimisation, modelling and simulation of the solutions of partial differential equations - Usage and development of 3D models of real objects - Integration of my code with an existing C++/MATLAB package
- Diplôme d'Ingénieur (Master's degree)École Centrale Paris2007Award: Hoschet Prize 2006 - 2007 (for exceptional academic achievement) • Acquired skills in physics (grade A+), modelling and simulation (grade A), computer science (grade A), continuum mechanics (grade A), fluid dynamics (grade A+) • Acquired commercial awareness skills alongside the science as part of the curriculum (e.g. management – grade A) • Worked in various environments during internships abroad (India, etc.)