Venetis-Paraskevas Pallikaras

McLean, VA, USAvenetispall@gmail.comLinkedInGitHub

Education

Georgia Institute of Technology

Atlanta, GA, USA

MSc in Analytics

Aug 2021 – Dec 2022

GPA: 4.0 / 4.0

Democritus University of Thrace

Xanthi, Xanthi, Greece

BSc & MSc in Electrical and Computer Engineering

Oct 2014 – Nov 2019

GPA: 8.85 / 10 "Excellent", 3.7 / 4.0 (WES Course by Course Evaluation)

  • Second Among Graduating Year (Salutatorian)
  • Thesis: Automated speaker diarization and recognition in videos (Grade: 10/10)

Skills

Programming Languages: Python, SQL, R, Java, JavaScript, C++, C

ML & DL Software: Tensorflow, Keras, Pytorch, Scikit-Learn, HDBSCAN, XGBoost, LightGBM, Optuna, HyperOpt, PySpark, PYG, TinyML

NLP & Generative AI Software: NLTK, SpaCy, BERTopic, KeyBert, Hugging Face, Transformers, vLLM, LangChain, LlamaIndex, Unsloth, Axolotl, Megatron, LangGraph, Azure AI Foundry, n8n, Agent Reinforcement Trainer (ART), Copilot Studio, AutoGen

Analytical Techniques: Machine Learning, Deep Learning, Web Scraping, Computer Vision, GraphML, NLP, LLMs, Generative AI, Reinforcement Learning, LLM-Finetuning, GRPO

Visualization: Tableau, Plotly, Dash, Seaborn, D3.js, Matplotlib

Data & Cloud Platforms: Databricks, Azure, AWS

Other: Git, Linux, HTML, CSS, Beautiful Soup, Selenium, Pinecone, MLflow, Weights & Biases, FastAPI, LangSmith, Modal, Vercel

Experience

Navy Federal Credit Union | Senior Data Scientist | Vienna, VA, USA

Sept 2023 – Present

Navy Federal Credit Union | Data Scientist | Vienna, VA, USA

Jan 2023 – Sept 2023
  • Designed and deployed a GenAI Insight Discovery Pipeline on Databricks for large-scale call and chat transcripts, performing record-level theme extraction (vLLM), recursive summarization (LangChain), and topic quantification. Converting unstructured member conversations into structured insights to enable early issue detection and data-driven decision-making.
  • Accelerated enterprise GenAI adoption by building reusable LLM notebooks, prompt templates, and starter workflows. This streamlined internal LLM access and reduced onboarding time for analysts and data scientists.
  • Created multiple NLP solutions, developing sentiment analysis, emotion classification, praise/complaint detection, a multi-stage aspect-based sentiment & summarization system that identifies the core topics/issues, maps it to the relevant product/feature, evaluates sentiment, and generates summaries per topic/issue. These models are used across business units to elevate member experience and operational efficiency.
  • Developed a RAG pipeline to assist our call representatives and other internal teams to automate their daily workflow. Built internal RAG evaluation system (LLM-as-a-judge) which was aligned with human scoring.
  • Delivered insight-driven reporting for multiple enterprise initiatives, including government shutdown impact analysis, competitive banking product comparison, CD product redesign support, loan application journey pain-point extraction, and OMNI (mobile/online banking) post-launch issue discovery.
  • Presented monthly at Databricks User Group, sharing internal NLP/GenAI solutions and advanced Databricks practices.
  • Developed predictive models, including a 6-month forward-looking member engagement model supporting proactive outreach and retention initiative by utilizing XGBoost and HyperOpt.

General Electric | Data Science Intern | Atlanta, GA, USA

May 2022 – Aug 2022
  • Explored model compression techniques and deployed deep learning models on Arduino using TensorFlow Lite for edge inference.

National Centre for Scientific Research "Demokritos" | Research Intern | Athens, Greece

July 2018 – Sept 2018
  • Analyzed the analog behavior of Memristors and executed Retention and Electrical Characterization measurements on them.

Projects & Publications

Predict likelihood of members having housing insecurity issues

Top 1 Accuracy | Sponsor: Humana — Fall 2022

  • Created an ensemble of decision tree-based algorithms using XGBoost, LightGBM and Optuna for hyperparameter tuning.

Published Paper: "Retrieval Augmented Generation for Liquid Sodium Facility Documentation Processing" — Dec 2024

Published Paper: "Agentic Retrieval Augmented Generation for Advanced Reactor Thermal Hydraulic System" — Dec 2024

Published Paper: "Artificial Neural Network Performance Boost using Probabilistic Recovery with Fast Cascade Training" — Nov 2020

Courses & Certifications

Certifications: AWS-CCP

Stanford AI Professional Program: ML with Graphs, NLP with DL, Natural Language Understanding (NLU)

Databricks: Generative AI fundamentals, Data Engineer Associate, ML Data Scientist Associate, Generative AI Engineer Associate