Lead Data Scientist & AI Systems Architect

I ship AI that
actually works.

PhD in Cognitive & Computer Science. 10+ years taking ML from research to production. 15 products shipped across ML, SaaS, iOS, and quantitative trading — including 4 from ideation to production in just 12 months.

PyTorch LLM Architecture Predictive Models RAG & Agents Full-Stack AI Quantitative Systems
Get in Touch Cognitive Intelligence Dev →

About

Research to
production. Every time.

Lead Data Scientist and AI Systems Architect with a PhD in Cognitive & Computer Science. I take ML projects from research through architecture, training, and deployment using PyTorch, TensorFlow, and scikit-learn on AWS/GCP and containerized infrastructure.

15 products shipped across my career. As Founder & CEO of Cognitive Intelligence Development, I shipped 4 from ideation to production in under 12 months — a sales intelligence SaaS, two iOS apps with LLM backends, and an automated quantitative trading system. Before that, I built predictive models at Stardust that hit 90%+ accuracy on women's health predictions, and led emotion recognition ML at Genoemote.

10+
Years in AI/ML
15
Products Shipped
PhD
Cognitive & CS
90%+
Model Accuracy

ML & Deep Learning

PyTorch, TensorFlow, scikit-learn, XGBoost

LLM Architecture & Agents

RAG pipelines, fine-tuning, prompt engineering

Predictive Modeling

Time-series, health predictions, behavioral scoring

Cloud & MLOps

AWS SageMaker, GCP Vertex AI, Docker, MLflow

Quantitative Systems

Monte Carlo, Kelly Criterion, real-time data

Cognitive & Behavioral Science

Human-centered AI, emotion recognition, fMRI/EEG

Technical Skills

The full stack.

ML & Deep Learning

PyTorch TensorFlow/Keras scikit-learn XGBoost Hugging Face NumPy Monte Carlo

LLM & GenAI

RAG Pipelines LoRA/QLoRA LLM Agents Prompt Engineering Claude API OpenAI API

Backend & Data

Python FastAPI SQLAlchemy (async) PostgreSQL Redis Pandas SQL R

Cloud & MLOps

AWS SageMaker EC2 / S3 GCP Vertex AI BigQuery Docker Fly.io Railway MLflow W&B

Infrastructure

Pinecone FAISS ChromaDB Stripe Firebase arq Queues WebSocket SSE

Mobile & Frontend

Swift SwiftUI SwiftData StoreKit 2 Slack Bolt Telegram Bot

Experience

From luxury brands to
production ML.

Mar 2025 — Present
Perpignan, France

Founder & CEO

Cognitive Intelligence Development

Founded AI/ML studio. Architected and shipped 4 production AI products in under 12 months across SaaS, iOS, and quantitative trading.

  • Prism — Sales conversation intelligence SaaS (FastAPI, PostgreSQL, Redis, async arq queue) analyzing call transcripts via Claude LLM pipelines with cognitive science frameworks. 0–100 scoring, deal risk estimation, win probability. Multi-tenant with Stripe billing, Slack/Telegram bots, embeddable widget.
  • AIM & TinyTaste — Two iOS apps (Swift/SwiftUI, SwiftData, Firebase, StoreKit 2) with FastAPI + Claude backends. AI-generated goal plans, safety validation, prompt injection detection, structured output parsing. Freemium subscription models.
  • AI Trading Bot — Fully automated quantitative trading system (Python, NumPy) for prediction markets. 5+ simultaneous strategies, real-time price feeds via WebSocket, Monte Carlo probability estimation (3,000 paths), Kelly Criterion sizing, whale wallet replication, AI-powered news impact analysis across 14 RSS feeds.
  • Common stack: Anthropic Claude API, FastAPI, Docker (Railway/Fly.io), PostgreSQL, Redis, security-first design (rate limiting, input sanitization, prompt injection detection).
LLM PipelinesFastAPIPostgreSQLRedisSwift/SwiftUIDockerMonte CarloWebSocket
Jan 2024 — Mar 2026
Remote

Head of Science and AI

Stardust App

Led data science and ML engineering for a women's health platform, owning the full model lifecycle from feature engineering to production serving.

  • Designed predictive models for cycle length, period length, and ovulation day using basal body temperature (BBT) time-series data — 80%+ real-time accuracy, 90%+ with optimal tracking.
  • Architected feature pipelines and model training strategy using gradient-boosted (XGBoost) and deep learning models (PyTorch) on physiological time-series data.
  • Architected a health correlations framework across full data frames of logs and wearable data.
  • Designed A/B testing frameworks and statistical validation pipelines to measure model performance against clinical benchmarks.
PyTorchXGBoostPandasFemTechTime-SeriesA/B Testing
2022 — 2023
Remote

Senior Data & Behavioral Scientist

CoachHub
  • Designed predictive analytics pipelines (scikit-learn, SQL/PostgreSQL) to quantify coaching outcomes across 3,500+ enterprise users.
  • Designed behavioral scoring models combining NLP-based sentiment analysis with engagement metrics to predict coaching effectiveness.
scikit-learnNLPPostgreSQLBehavioral ScienceEnterprise
Dec 2021 — 2022
Remote

Interim Chief Technology Officer

Genoemote
  • Led architecture and development of emotion recognition ML system — real-time classification pipeline processing audio and facial expression data.
  • Managed model training infrastructure on AWS (EC2/S3) and implemented CI/CD for model deployment.
Emotion AIAWS EC2/S3CI/CDReal-Time MLMultimodal
Nov 2020 — 2023
Miami, FL

Lead Cognitive & Computer Scientist, Co-Founder

Humaine AI, LLC
  • Co-founded AI consultancy; designed and deployed ML solutions for clients in healthcare, education, and consumer tech.
  • Built human-data modeling frameworks integrating cognitive science with supervised/unsupervised learning approaches.
AI ConsultingHealthcare MLCo-Founder
2018 — 2019
Treviso, Italy

Advanced Technologies Lead Researcher & PM

H-Farm
  • Led R&D and delivery of VR/AR systems for luxury brands (Bulgari, Gucci, Ferrari) and energy sector (Enel, SNAM).
  • Designed data collection pipelines for user interaction analytics in immersive environments; informed product iteration with quantitative user research.
VR/ARLuxury BrandsUser AnalyticsR&D
2016 — 2017
Izmir, Turkey

UX Designer & Product Manager

Ege Bilgi
  • Led user research and data-driven design for web and mobile products using A/B testing and behavioral analytics.
UX ResearchA/B TestingProduct Management

Builds

Products I've shipped.

15 products shipped across my career. 4 built from ideation to production in 12 months at CID alone — SaaS, iOS, and quantitative trading.

💡

Prism

SaaS Platform

Sales conversation intelligence powered by cognitive science. Analyzes call transcripts via Claude LLM pipelines, generates 0–100 scoring with deal risk estimation and win probability. Multi-tenant, async job queue, Slack/Telegram bots.

FastAPI · PostgreSQL · Redis · arq · Claude API · Stripe · Docker
cogdev.ai →
🧠

AIM

iOS App

AI-powered goal achievement. Users set goals, AI generates personalized science-backed milestone plans with subtasks. Conversational creation flow, safety validation, prompt injection detection, structured output parsing.

Swift/SwiftUI · SwiftData · Firebase · StoreKit 2 · FastAPI · Claude API · SSE
cogdev.ai/aim →
🍽

TinyTaste

iOS App

Baby weaning & nutrition intelligence. Tracks food introduction with reaction scoring and allergen flagging. AI-powered personalized insights, recipe generation, and seasonal produce suggestions. Multi-baby profiles.

Swift/SwiftUI · SwiftData · Firebase · StoreKit 2 · FastAPI · Claude API · Redis
cogdev.ai/tinytaste →
📈

AI Trading Bot

Quantitative System

Fully automated prediction market trading system running 5+ simultaneous strategies. Real-time price feeds via WebSocket, Monte Carlo probability estimation (3,000 paths), Kelly Criterion sizing, whale wallet replication, AI news impact analysis.

Python · NumPy · WebSocket · Claude API · Polymarket CLOB · Binance API · Telegram

Education

Trained across
four countries.

PhD Brain, Mind and Computer Science
University of Padova & University of Edinburgh
2016 — 2019
Research on emotion recognition in AR/VR using ML classification models and physiological signal processing
MSc Cognitive Systems and Interactive Media
Pompeu Fabra University, Barcelona
2014 — 2015
Quantitative research methods; fMRI/EEG signal analysis and statistical modeling
BA (Hons) Media & Entertainment Technology
University of Portsmouth
2011 — 2014
Foundation in interactive technology, media systems, and human-computer interaction
🇬🇧
English
Fluent
🇹🇷
Turkish
Native
🇪🇸
Spanish
Professional
🇮🇹
Italian
Conversational

Contact

Let's build something.

Whether it's AI consulting, ML engineering, research collaboration, or building something new — I'm always open to conversations.

✉ bnndmr@gmail.com ◆ cogdev.ai ☎ +33 7 88 51 52 08