Sarah LAZRI

Dorthia is a precision regulatory platform that transforms compliance into a competitive advantage for regulated industries (biotech, pharma, cosmetics, healthcare, pharmacies). We combine AI Act readiness, GDPR expertise, and adaptive learning to help companies get audit-ready. Our two proprietary modules — Dorthia Compliance (adaptive learning for general compliance) and Dorthia AI Act (tracked AI literacy per profile) — automate documentation, watermark AI outputs, track employee training, and prepare clients for notified body certification. Key Competencies : AI Act & GDPR expertise – strategic and operational mastery of both regulations Adaptive learning design – personalized compliance training that adjusts to each employee's pace Tracked AI literacy module – per-profile monitoring, scoring, and notifications for AI Act compliance (Art. 4) Automation & no-code workflows – API integrations with Airtable, Slack, Teams, and ERP systems Data governance & bias auditing – automated dataset scanning and bias detection Machine-readable watermarking – invisible watermarking for images, audio, and text (Art. 50) Gap analysis & audit simulation – white audits, risk calculators, and notified body preparation 20-week certification roadmap – step-by-step client journey from audit to certification Packaged offers & partnerships – Essential, Certification, Premium plans + white-label for law firms Continuous regulatory watch – annual subscription with automatic template and module updates

CognMap is an advanced Digital Brain Twin platform designed to integrate seamlessly into hospital infrastructures. It provides clinicians with a powerful tool for early diagnosis, therapeutic personalization, and clinical simulation. Unlike traditional High-Performance Computing (HPC) systems, CognMap operates on a proprietary operating system that activates only necessary neural circuits, making it significantly more energy-efficient, economical, and accessible for daily medical use. Our system maps over 1,200 neurological functions, 60 activity types, and 1,400+ brain regions, with dynamic modeling of connections, activity gradients, and pathological interactions. CognMap represents a new frontier in neurology, bridging data, discovery, and personalized care. A new era in brain health and neuroscience.

PharmAlgo is an innovative company specializing in the intersection of advanced mathematics, algorithmic computing and biopharmacy. Its main objective is to revolutionize pharmaceutical research and personalized medicine by offering tailor-made solutions ranging from genomic analysis to predictive drug modeling. 1. Genomic Analysis and Biomarkers PharmAlgo develops advanced algorithms to analyze genomic data and identify relevant biomarkers. This enables : A better understanding of diseases (cancers, rare diseases, etc.). Classification of patients according to their genetic profile for more personalized medicine. Detection of genetic mutations linked to response to treatment. 2. Optimizing clinical trials The company uses mathematical and computational models to improve the design and management of clinical trials: Reduce costs and delays by identifying the best protocols. More efficient patient selection using algorithmic criteria. Predictive analysis of results to adjust trials in real time. 3. Predictive drug modeling PharmAlgo integrates numerical simulation and AI techniques to : Predict molecule efficacy and toxicity prior to in vitro testing. Accelerate drug discovery by reducing empirical trials. Personalize treatments according to the patient's profile.

Dans le cadre de cette mission, j’ai occupé un rôle d’architecte de la donnée, avec pour objectif de structurer et rendre interopérables des écosystèmes complexes de données de santé. J’ai été initialement sollicité en raison de mes travaux sur des approches avancées de structuration et de mise en relation des données (notamment autour de concepts de “bridge” de données à haute complexité), visant à faciliter l’intégration de sources hétérogènes. Mes principales responsabilités ont ensuite porté sur : La conception et la structuration de modèles de données adaptés à des projets d’agents IA en santé Le soutien à des projets de recherche nécessitant une normalisation et une interopérabilité forte (notamment via SNOMED CT) L’analyse et la préparation des infrastructures de données dans le cadre de réflexions autour d’une future faculté de médecine “AI-friendly”, intégrant nativement les usages de l’intelligence artificielle Cette expérience m’a permis d’intervenir à l’interface entre expertise métier santé, ingénierie des données et innovation en intelligence artificielle.