Felipe A. Moreno-Vera

Data Scientist | AI Researcher | PhD Computer Science

Currently, I am working on Multimodal Retrieval-Augmented Generation (RAG) systems and Vision-Language Models (VLMs), with a strong focus on advancing multimodal large language models (MLLMs) and Explainable AI. My research explores how visual and textual information can be effectively integrated to build intelligent systems that are both accurate and interpretable.

In particular, I am interested in developing efficient and robust models for analyzing and understanding human perceptions from street view imagery. This includes investigating how urban environments are perceived in terms of safety, liveliness, aesthetics, accessibility, and overall quality of life. By extracting meaningful perceptual and semantic features from large-scale visual data, my work aims to support applications in urban computing, smart cities, and data-driven urban planning.

More broadly, I am motivated by the challenge of creating multimodal AI systems that can reason across different data modalities while remaining transparent, explainable, and adaptable to real-world scenarios.

Employment

Agentic-AI Engineer June 2024 - Current
GenVER Applied Intelligence (GenVER-AI) Remote 🌎

• 🇵🇾 COOMECIPAR
  – CooRAG: RAG-based chatbot to retrieve, summarize, and explain relevant financial and regulatory information from Paraguayan credit cooperative documents.
  Technologies: RAG-Ops, Huggingface, Claude, OpenAI, LDAP
  – CooGPT: LLM-based chatbot to upload, understand, and answer queries about relevant financial and regulatory information from Paraguayan credit cooperative documents.
  Technologies: AWS, LLM-Ops, FastAPI, Gemini, OpenAI

• 🇧🇷 FGV
  – LegalXplain: Apply Large Language Models (LLMs) methods to explain relevant binding references in legal documents from the Brazilian Supreme Court (STF).
  Technologies: Azure, RAG, LangChain, Langgraph, OpenAI, Legal-NLP

• 🇵🇪 ONPE
  – SEActas: Unifying voting records from different scanner devices using Efficient OCR-Nets for signatures and votes recognition for Peruvian general elections.
  Technologies: ML-Ops, FastAPI, OCR-nets, Large language models, Min.io

• 🇨🇱 FALABELLA
  – BoTopicos: Agentic RAG-based bot to identify and categorize user opinions using semantic and keyword matching to associate and classify new user opinions.
  Technologies: RAG-Ops, Perplexity, Spacy, NLP, LLM, Hierarchical classification

AI Engineer June 2022 - June 2024
Getter Amplified Industry (GetterAI) Manaus - Brazil 🇧🇷

• Samarco: Monitor the entire industrial operation process and generate audible or visual alerts when a risk is detected. The development of end-to-end applications focuses on the detection of safety implements and the correct analysis of potential risks in industries.

• Akzonobel: Building an infrastructure to infer safety risks and protection regulations. Implementing AI end-to-end applications to assess posture and ergonomics, detect work cycles and times elapsed, and determine factory safety areas.

Technologies: AWS, Kuberflow, Docker, Kubernetes, MLOps

AI Data Scientist May 2019 - May 2022
Fundação Getúlio Vargas (FGV) Rio de Janeiro - Brazil 🇧🇷

• LegalAnalytics: Exploring and predicting possible citations between Brazilian legal documents. Analysis, processing, organization, and extraction of the content of legal documents from the Brazilian Supreme Court-STF. Through NLP and LLMs, it is possible to identify and predict the most relevant sentences.

• TimberFlow: Detection of deforestation and illegal transport in the Amazon. Integration of Sinaflor and Sisflora, Brazil’s two most relevant databases on timber transport. Through Markov chains and graph networks, it is possible to predict the most likely transport flow between a consumer and a producer.

• UrbVis: Analysis of criminal data and correlations with urban security perception. Through convolutional networks, segmentation, and object detection, it is possible to identify, process and extract the main characteristics of street images in order to predict the perception of security in the cities of Rio de Janeiro and Sao Paulo.

Technologies: Azure, TensorFlow, Keras, PyTorch, Computer Vision, NLP

Web/Mobile Developer August 2017 - October 2018
CERNICALO S.A. Lima - Perú 🇵🇪

• TotemApp: Android-based mobile application to be installed on ONPE’s Totem devices, aimed at informing and responding to inquiries from Peruvian citizens using Android SDK versions from Jelly Bean (API 18) to Marshmallow (API 23).

Technologies: Android, Java, PHP, Laravel, MySQL, Apache Kafka.

Research Assistant January 2016 - July 2017
Centro de Tecnologías de Información y Comunicaciones (CTIC) Lima - Perú 🇵🇪

• BeaGOns: Real-Time Fog-Edge web application to read and process sensor data. Using an edge computing architecture, we deploy Raspberry Pi 2B-based modules (Edge Layer) across Lima City to collect environmental data via sensors.

• CHESS: High-Performance Beowulf Hybrid CPU-GPU Cluster with NVIDIA GeForce GTX 980 Ti. Using Apache Spark, MPI, and OpenMP, we simulate urban transport traffic with 1.2TB of temporal data from New York City and Seoul.

Technologies applied: C/C++, HPC, Fog Computing, Golang, Embedded systems

Education

• Fundação Getúlio Vargas (FGV) Rio de Janeiro - Brasil 🇧🇷
  Ph.D. in Computer Science 2022 - 2026
  Thesis: Exploração da percepção de segurança urbana com modelos de visão-linguagem e contrafactuais - Slides
  Advisor: Ph.D. Jorge Poco
  Contributions:
  – Journal From Vision-centric to Multimodal: A Systematic Review of Street View–Based Urban Perception, TETCI 2026
  – Conference Assessing Urban Environments with Vision-Language Models: A Comparative Analysis of AI-Generated Ratings and Human Volunteer Evaluations, IJCNN 2025
  – Conference UrbanPhysicalDisorder-4K: Understanding Urban Perception via Counterfactuals and Street View Signs of Physical Disorder, BigData 2025
  – Conference What Makes a Place Feel Safe? Analyzing Street View Images to Identify Relevant Visual Elements, WI-IAT 2024

• Universidade Federal do Rio de Janeiro (UFRJ) Rio de Janeiro - Brasil 🇧🇷
  M.Sc. in Computer Science 2021 - 2023
  Thesis: Detecção Automatizada de Exploração de Vulnerabilidades em Fóruns de Hacking Clandestinos - Slides
  Advisor: Ph.D. Daniel Menasché
  Contributions:
  – Book-chapter Beneath the Cream: Unveiling Relevant Information Points from CrimeBB Underground Forums with Its Ground Truth Labels, CSCML 2024
  – Conference Cream Skimming the Underground: Identifying Relevant Information Points from Online Forums, CSR, 2023
  – Conference Inferring Discussion Topics about Exploitation of Vulnerabilities from Underground Hacking Forums, ICTC, 2023

• Universidad Católica San Pablo (UCSP) Arequipa - Perú 🇵🇪
  M.Sc. in Computer Science 2018 - 2020
  Thesis: Técnicas de Aprendizaje Profundo para el Análisis de la Percepción de la Seguridad Urbana - Slides
  Advisor: Ph.D. Jorge Poco
  Contributions:
  – Conference Quantifying Urban Safety Perception on Street View Images, WI-IAT 2021
  – Book-chapter Urban Perception: Can We Understand Why a Street Is Safe?, MICAI 2021
  – Book-chapter Understanding safety based on urban perception, ICIC 2021

• Universidad Nacional de Ingeniería (UNI) Lima - Perú 🇵🇪
  B.Sc. in Computer Science 2012 - 2017
  Thesis: - Slides
  Advisor: Ph.D. Glen Rodriguez

  Final Project: Diseño e implementación del core level de una plataforma transversal basado en arquitecturas Fog Computing - Slides
  Advisor: Ph.D. Manuel Castillo

Certifications and Specializations

• Data Engineering, Big Data, and Machine Learning on GCP Specialization.
• Natural Language Processing Specialization.
• Deep Learning Specialization.
• Machine Learning Specialization.
• Mathematics for Machine Learning and Data Science Specialization.
• AI4Medicine Specialization
• AI4Good Specialization.

Honors and Awards

• 2026 Attendant Travel Grant Award: Granted by SALA AI.
• 2025 Innovative Display of Daily Communication Patterns Award: Granted by IEEE VAST Challenge.
• 2025 Student Travel Grant Award: Granted by International Neural Networks Society.
• 2024 Strong Data Enrichment and Augmentation Award: Granted by IEEE VAST Challenge.
• 2023 Best Urban Mobility Solution Award: Granted by UFRJ and FAPERJ.
• 2022 Brazilian Doctorate Scholarship: Granted by FAPERJ and CAPES.
• 2021 Brazilian Master Scholarship: Granted by FAPERJ and CNPq.
• 2019 Outstanding Poster Award: Granted by ICML-LXAI Workshop.
• 2018 Peruvian Master Scholarship: Granted by Ciencia Activa, CONCYTEC, and FONDECYT.
• 2017 Higher Fifth: Higher Tenth computer science class 2017.
• 2016 Outstanding Students Award: Granted by Universidad Nacional de Ingenieria (UNI).
• 2015 Peruvian Bachelor Scholarship: Granted by PRONABEC.

Latest Publications

To know further about my research line, visit this link.

Volunteer Work

• Social program volunteer like Crea+ and Techo-Peru.
• Innovation Mentor at the technological incubator StartUp UNI, Lima, Perú.
• Member of the University Assembly from 2016 to 2017.
• Member of Student Center from 2012 to 2013.

Additional Skills

Category	Proficiency in approximate descending order from left to right
Programming Languages	C, C++, Go, Python, R, M (Octave/MATLAB), Javascript
Web Technologies	HTML, CSS/SCSS, Django, VueJS, ReactJS, Flask, BeeGo, Gorilla, Node.js, Angular, Jekyll
Databases/Storage	PostgreSQL, MySQL, MongoDB
Cloud	AWS, GCP
Productivity Tools	LaTeX, GIT, Jupyter
Software Engineering	Test-Driven Development: Selenium

References

Available on request.