Péter Róbert

Péter Róbert

Founding Engineer | AI Systems Architect

Full-Stack AI Developer with expertise in deep learning, distributed systems, and frontend architecture. Specialized in shipping end-to-end AI products from multi-agent orchestration to polished user interfaces.

About Me

I'm a Full-Stack AI Developer and Founding Engineer with a strong foundation in deep learning, distributed systems, and frontend architecture. I have extensive experience in shipping end-to-end AI products, from architecting multi-agent orchestration engines using LangGraph to building polished, responsive interfaces with Next.js and React.

My journey spans from developing AI-powered sales automation platforms at Salestools, where I operate as a hands-on technical architect, to building high-performance frontend applications for AI-driven image generation platforms. I have a proven track record of designing scalable microservices and implementing complex business logic like rotational messaging systems and real-time credit economies.

Currently pursuing a Master's in Artificial Intelligence at UMFST "George Emil Palade" in Târgu Mureș, focusing on U-Net architectures for automatic lung segmentation in chest X-rays. I'm passionate about bridging the gap between cutting-edge AI research and production-ready applications.

Work Experience

Founding Engineer

SalestoolsBucharestJun 2025 - Present

  • Engineered a robust, distributed outbound mail sending system with rotational email logic for high-volume campaigns
  • Designed and implemented a transactional credit deduction system with real-time balance tracking
  • Architected a LangGraph-based AI orchestrator for intelligent multi-agent scheduling and timezone coordination
  • Built FastAPI-based microservices handling async workflows with SQLAlchemy and PostgreSQL
  • Developed core UI components for complex state management using React and TypeScript
  • Built scalable integration layer connecting Google Workspace, Microsoft 365, HubSpot, and Salesforce with Redis caching

Frontend Developer

Tengr.aiRemoteNov 2023 - Mar 2025

  • Built responsive AI-powered interfaces using React, Next.js, and TypeScript
  • Integrated Pixi.js for high-performance, GPU-accelerated graphics rendering
  • Managed cloud infrastructure and deployments via Vercel and AWS S3
  • Implemented secure authentication flows using Auth0 and optimized application performance

Assistant Analyst

AccentureTârgu MureșJun 2023 - Aug 2023

  • Developed forecasting tools using ARIMA models and Python for industrial data analysis
  • Integrated external engines with OSIsoft PI Systems for advanced visual analytics

Web Developer Intern

ReeaTârgu MureșJun 2022 - Aug 2022

  • Developed a fully functional e-commerce webshop using Ruby on Rails and SQL
  • Optimized database queries for performance and implemented core e-commerce features

Featured Projects

Exocortex - AI Second Brain

Exocortex - AI Second Brain

2025

Cross-platform personal AI assistant that captures voice notes, transcribes them using Groq Whisper, stores them with pgvector semantic search, and enables natural conversation with your memories. Built with Flutter, Supabase, and multi-agent AI orchestration.

FlutterSupabaseGroq AIpgvectorLLM AgentsVoice Recognition
AI Agent Calendar System

AI Agent Calendar System

2025

Autonomous scheduling system using reasoning loops and error recovery logic to manage calendar conflicts without human intervention. Built for Salestools internal operations.

LangGraphPythonAI AgentsFastAPI
U-Net 3D Lite for Lung Segmentation

U-Net 3D Lite for Lung Segmentation

2025

Reduced model parameters by 73% (15M to 4M) and VRAM usage by 75% while maintaining a Dice score of 0.9448. Designed to run efficiently on consumer hardware (RTX 4070) for real-world medical constraints.

PyTorchU-NetComputer VisionMedical AI
X-Ray Classification Pipeline

X-Ray Classification Pipeline

2024

Combined U-Net-based lung segmentation with classification models using Grad-CAM and transfer learning for explainable AI diagnostics. Applied ROC-AUC and precision-recall metrics.

TensorFlowGrad-CAMTransfer LearningMedical Imaging
Industrial Automation VR Simulation

Industrial Automation VR Simulation

2024

Independently designed and built a fully functional 3D VR simulation of a Soda Factory using Visual Components. Demonstrated real-time interactivity and automation.

VRVisual Components3D ModelingIndustrial Automation
Stratego Game AI

Stratego Game AI

2023

Developed a decision-making algorithm for environments with incomplete information. Implemented heuristic search techniques to handle uncertainty and optimize competitive gameplay strategies.

PythonGame AIHeuristic SearchAlgorithm Design
E-Commerce Webshop

E-Commerce Webshop

2022

Full-featured e-commerce platform built with Ruby on Rails and SQL. Implemented secure payment processing, inventory management, and optimized database queries.

Ruby on RailsSQLE-CommercePostgreSQL

Skills & Technologies

Languages

  • Python
  • TypeScript
  • JavaScript
  • C++
  • SQL
  • Ruby

Frontend

  • React
  • Next.js
  • Tailwind CSS
  • Pixi.js
  • Vite
  • Zustand

Backend & Cloud

  • FastAPI
  • PostgreSQL
  • Redis
  • Docker
  • AWS S3
  • Vercel
  • SQLAlchemy

AI & ML

  • PyTorch
  • TensorFlow
  • LangGraph
  • LangChain
  • DSPy
  • OpenAI/Anthropic APIs
  • Computer Vision

Concepts & Tools

  • Microservices
  • CI/CD
  • Multi-Agent Systems
  • RESTful APIs
  • Agile
  • Auth0

Education

Master's in Artificial Intelligence

UMFST "George Emil Palade" • Târgu Mureș • 2023 - 2025

Thesis: Comparative Analysis of U-Net Architectures for the Automatic Segmentation of Lungs in Chest X-Rays

Bachelor's Degree in Informatics

UMFST "George Emil Palade" • Târgu Mureș • 2020 - 2023

Thesis: Strategy optimization for incomplete information (Stratego AI)

Get In Touch

I'm currently open to new opportunities and interesting projects. Whether you have a question or just want to say hi, feel free to reach out!