A

Andrei Eduard Lupica

About

Detail

Odense, Denmark
Contact Andrei regarding: 
Flexible work
Starting at USD20/hour
id_card
Internships
Starting at USD1.8k/month

Timeline

work
Job
school
Education
folder
Project
flag
Award
auto_stories
Publication

Résumé

Jobs verified_user 0% verified
  • London Stock Exchange Group (LSEG)
    Ai Engineer public Remote experience
    London Stock Exchange Group (LSEG)
    May 2026
    • Built an LLM-driven pipeline in Python that validates corporate ownership structures via web search and retrieval-augmented generation, automatically
    detecting entity changes such as mergers, acquisitions, bankruptcies, and rebranding, and providing supporting evidence to streamline compliance
    reviews
    • Designed and executed an evaluation framework combining automated metrics and human review to measure accuracy, reliability, and cost, while
    identifying key failure modes
    • Delivered a production-oriented prototype, stakeholder presentations, and a technical report outlining scalability constraints, cost trade-offs, and
    deployment recommendations
  • CloudGeometry
    AI & MLOps Engineer
    CloudGeometry
    Oct 2025 - Current (1 year 1 month)
    • Engineered scalable Retrieval-Augmented Generation (RAG) data pipelines with sub-second latency using AWS Bedrock, ECS
    containers, and AuroraDB for text chunking, embedding, and vector search, incorporating data preprocessing workflows to enhance
    model input quality.
    • Architected multi-stage AI agent workflows that reduced manual data evaluation time by 60% through chained LLM prompts, external
    API validation, dynamic scoring logic, and conversational automation, with rigorous model evaluation to ensure response accuracy.
    • Led the architectural design and orchestration of a centralized low-code AI platform that enabled the rapid deployment of 10+ custom
    NLP-powered agents via Langflow, AWS Secrets Manager,
  • Deutsche Telekom
    Data Scientist
    Deutsche Telekom
    Aug 2025 - Current (1 year 3 months)
    • Developed enterprise data ingestion and processing pipelines with 99.9% uptime using Dagster to orchestrate chunking, sorting, and
    indexing of raw unstructured data into Elasticsearch and MongoDB, enabling downstream machine learning workloads.
    • Designed a custom context-aware RAG architecture linking vectorized Elasticsearch data with the OpenAI GPT API, leveraging
    predictive analytics to anticipate user query intent and increase resolution speed by 35%.
    • Programmed the dynamic generation of specialized conversational AI interfaces from a unified data source using Python and
    LLMOps, employing adaptable LLM prompts and advanced retrieval logic, which streamlined development and improved response
    relevance f
  • Digital Nexus AI
    Generative AI Engineer (Internship)
    Digital Nexus AI
    Mar 2025 - Aug 2025 (6 months)
    • Co-developed a Software Engineer AI Agent that converted feature tickets into validated pull requests, reducing average PR
    turnaround time by 42%.
    • Improved automated code quality by increasing test coverage by 25% through LLM-driven code suggestions and validation steps.
    • Authored an evaluation harness with pytest and GitHub Actions that benchmarked generated outputs against coding standards,
    ensuring every submission met functional requirements before merge
  • R
    AI/ML & Robotics Engineer
    RoboAl
    Sep 2024 - Apr 2025 (8 months)
    • Automated a generative AI pipeline for drafting legal statements that reduced manual document processing time by over 70% using
    the Google Gemini API, ChromaDB, and Python-based structured parsing.
    • Built a mobile remote surveillance system that integrated thermal and 360-degree cameras onto a Unitree Go2 robot over a secure
    VPN, enabling continuous remote monitoring and reducing on-site inspection trips
    • Created highly precise 3D surface mapping of geological samples by developing C++ and Python ROS nodes to coordinate a UR5
    robotic arm with a Gocator line scanner.
Education verified_user 0% verified
  • S
    B.Eng., Artificial Intelligence & Data Engineering
    Satakunta University of Applied Sciences
    Sep 2023 - Feb 2026 (2 years 6 months)
    Pori
Projects (professional or personal) verified_user 0% verified
  • Ultrahack
    Defense Innovation Network Finland Hackathon (Powered by Sako) 2025 Winner
    Ultrahack
    May 2025
    Real-Time Wind Prediction Model Secured 1st place (€4,000 grand prize) out of many international teams with Team O21 at the DEFINE Hackathon powered by Sako. Co-developed a camera-only, real-time downrange wind estimation system for bullet trajectories. Collaborated to engineer the end-to-end pipeline using ResNet-18 for per-frame feature extraction and an LSTM for temporal drift modeling, delivering a fully validated demo in under 48hours.
Awards verified_user 0% verified
  • Ultrahack
    1st Place Winner | Ultrahack DEFINE Hackathon powered by Sako 2025
    Ultrahack
    May 2025
    Real-Time Wind Prediction Model Secured 1st place (€4,000 grand prize) out of many international teams with Team O21 at the DEFINE Hackathon powered by Sako. Co-developed a camera-only, real-time downrange wind estimation system for bullet trajectories. Collaborated to engineer the end-to-end pipeline using ResNet-18 for per-frame feature extraction and an LSTM for temporal drift modeling, delivering a fully validated demo in under 48hours.
Publications verified_user 0% verified
  • S
    Minimal command AAC: Comparing Classic Signal Processing vs AI on Minimal EEG/EMG
    Satakunta University of Applied Sciences - SAMK
    Sep 2025 - Feb 2026 (6 months)
    Designed and deployed a real-time brain-computer interface comparing rule-based signal processing against AI models (CNN, Random Forest) for minimal-command AAC. Handled end-to-end data collection, filtering, and model training.
    https://www.theseus.fi/handle/10024/909471