Leticia Lorena Rodríguez

About Me

Leticia Lorena Rodríguez, Professor in Computer Science, is the Founder of TEG AI Consulting and an Invited Professor at the University of Buenos Aires, dedicated to bridging the gap between academic research and enterprise innovation. Her journey in technology began early; coding by age 12 with dreams of becoming a software engineer, she launched her professional career at just 18 at PeopleSoft, straight out of high school.

Over the next 15 years, Leticia built a robust foundation in the corporate tech industry, including a tenure as a Senior Software Engineer and Team Leader at MercadoLibre. Driven by a growing fascination with computer science and education, she took a bold step: leaving the corporate world to join a research facility at CONICET as a Research Support Technician. During this transformative period, she began formalizing her academic expertise, eventually earning her degree as a Professor in Computer Science from the University of Buenos Aires (UBA).

In 2017, discovering Artificial Intelligence through online courses became the catalyst that completely redefined her trajectory. From that point on, her passions for scientific research, university-level teaching, and software engineering converged entirely around AI. This deep specialization propelled her to Google, where she served as an AI Consultant, advising C-level executives and leading AI engineering teams to deploy cutting-edge Generative AI solutions.

Today, Leticia continues to shape the future of tech and learning in LATAM through multiple avenues. Alongside her roles at the University of Buenos Aires and TEG AI Consulting—where she focuses on empowering organizations through specialized AI education, governance, and advisory services—she is deeply engaged in independent scientific research. Focusing on Explainable AI (XAI) and AI Safety, her independent work culminated in the publication of her paper, "Explaining is Harder Than Predicting Alone: Evaluating Concept-based Explanations of MLLMs as ICL Visual Classifiers," which she is proudly presenting at this conference.

Experience & Academic Roles

TEG AI Consulting

Founder & Strategic AI Advisor

Empowering organizations through specialized AI education, robust governance frameworks, and strategic advisory services. Driving digital transformation with state-of-the-art Generative AI solutions.

University of Buenos Aires (UBA)

Professor in Computer Science

Delivering high-level education in Computer Organization, Algorithms, Data Structures, and Artificial Intelligence. Dedicated to shaping the next generation of software engineers and AI researchers.

Teaching Projects

UBA - ECI 2026 AI Agents

Construyendo Agentes De Inteligencia Artificial

Los Agentes de IA y los Agentes Conversacionales están redefiniendo la interacción humano-computadora. Este curso intensivo ofrece una inmersión completa, desde los fundamentos teóricos hasta la implementación práctica, explorando el ciclo de vida completo del desarrollo de agentes y el diseño de la conversación.

Research Projects

Leticia conducts independent research to ensure AI is transparent, equitable, and designed to uplift underrepresented communities. A core area of action is Explainable AI (XAI). She collaborates with organizations to navigate the complexities of XAI, offering custom research tailored to specific business challenges to make AI systems more transparent and compliant.

Research Paper XAI

Explaining is Harder Than Predicting Alone: Evaluating Concept-based Explanations of MLLMs as ICL Visual Classifiers

In-context learning (ICL) enables multimodal large language models (MLLMs) to classify images from a few labelled examples. Yet, how these models use the provided context remains opaque. Evaluating four state-of-the-art MLLMs via an independent LLM-as-a-judge pipeline, we demonstrate that explaining is genuinely harder than predicting alone. Surprisingly, forcing models to generate formally structured, concept-based explanations degrades predictive accuracy monotonically, contradicting the assumption that explicit reasoning universally aids performance.

Read on arXiv →

Old Projects

Deep Learning

Retail Product Classification

Leveraged Deep Neural Networks and TensorFlow to classify grocery products based on shape and storage requirements, optimizing supply chain logistics.

Computer Vision

Real-Time Player Detection

Engineered a high-performance, real-time soccer player detection system operating over live video feeds, utilizing advanced Intel Architecture (SIMD) optimizations.

Reinforcement Learning

Autonomous Navigation via Q-Learning

Implemented a robust Q-Learning algorithm to solve complex gridworld navigation challenges, successfully guiding an agent while avoiding hazardous cliff regions.

Games - FPGA

FPGA Super Mario Bros

Using FPGA technologies, this repository builds a Super Marios Bros. level. The challenge is to use very little memory and sprites to recreate the classic mechanics.

ML Competition

Scene Recognition ECI 2019

Classification of scenes in images of accommodations. Machine Learning competition for ECI 2019 (UBA). Achieved a top-20 ranking and advanced to the next round.

Teaching Archive

Teaching Intro to AI (2018)

Artificial Intelligence Course

Introductory course of Artificial Intelligence prepared in 2018. Includes slides and labs for on-site learning covering RL (Monte-Carlo, Q-Learning), Deep Learning, CNNs, and RNNs.

Blog & Articles

Medium Article April 2018

Which Udacity's Nanodegree to take to learn AI, Deep Learning, or Machine Learning?

Are you thinking to start learning Artificial Intelligence or improve your existent knowledge? In this post, I share my opinion on Udacity's School of AI programs and hints that may help you decide the best Nanodegree for your skill set.

Read Article on Medium →

Medium Article Intelligence Engineering

The Rise of Intelligence Engineering: Deep Diving into Jensen Huang's Vision

An in-depth analysis exploring the emerging field of Intelligence Engineering. This article dives into Jensen Huang's strategic vision for the future of AI, discussing how engineering practices are evolving in the age of generative models and large-scale intelligence systems.