Jesús is available for hire
Hire JesúsJesús Martínez
Verified Expert in Engineering
Software Engineer and Developer
Bogotá - Bogota, Colombia
Toptal member since November 11, 2019
Bio
Jesús excels at creating high-quality software solutions due to his unique background, which combines 10+ years of top-notch software engineering with 8+ years of AI, machine learning, and data science work in areas like natural language processing (NLP) and computer vision, empowering him to be a valuable asset at both ends of the spectrum. He's worked with a wide array of tools and technologies, including Java, Python, TensorFlow, PyTorch, LLMs, RAGs, SQL, NoSQL, and DeepSeek.
Portfolio
DataSmarts
Scala, Java, TensorFlow, AutoML, Scikit-learn, NumPy, Pandas, PyTorch, Python...
Cortado, Inc.
Artificial Intelligence (AI), Retrieval-augmented Generation (RAG)...
Komorebi Investment Pty Ltd
Data Science, ETL, Databases, ETL Tools, Data Cleaning, Data Pipelines, Llama
Experience
- Python - 10 years
- Machine Learning - 10 years
- Natural Language Processing (NLP) - 5 years
- LangChain - 4 years
- TensorFlow - 3 years
- Deep Learning - 2 years
- Computer Vision - 2 years
- DeepSeek - 1 year
Availability
Part-time
Preferred Environment
Python, Linux, SQL, NoSQL, REST, OpenAI GPT-4 API, JavaScript, Large Language Models (LLMs), Retrieval-augmented Generation (RAG), Artificial Intelligence (AI)
The most amazing...
...thing I've built is a people counter and height detection algorithm based on computer vision techniques applied to RGB images and depth sensor data.
Work Experience
Founder | AI Consultant
2017 - PRESENT
DataSmarts
- Authored an Amazon best-selling book on computer vision recipes using TensorFlow 2 in partnership with Packt.
- Published weekly blog articles in Spanish, elucidating machine learning and computer vision concepts at both low and high levels.
- Produced weekly Spanish video tutorials on prominent computer vision and machine learning topics, featuring comprehensive step-by-step instructions and explanations.
- Developed and openly shared the source code for numerous computer vision projects, accompanied by comprehensive comments and instructions for execution and customization.
Technologies: Scala, Java, TensorFlow, AutoML, Scikit-learn, NumPy, Pandas, PyTorch, Python, Keras, Machine Learning, Artificial Intelligence (AI), ChatGPT, OpenAI GPT-4 API, OpenAI GPT-3 API, SQL, Language Models, Neural Networks, Databases, Llama
LLM Engineer
2024 - 2025
Cortado, Inc.
- Enhanced the state-of-the-art LLM-powered AI agent significantly for the property rental industry.
- Increased the efficiency of the data ingestion pipeline, resulting in lower costs and better user experience.
- Helped plan, design, and execute the migration from a monolithic architecture to a distributed event-based one.
Technologies: Artificial Intelligence (AI), Retrieval-augmented Generation (RAG), Large Language Models (LLMs), Machine Learning Operations (MLOps), LangChain, LlamaIndex, Vector Databases, Amazon SageMaker, DSPy, Real Estate, Llama
Senior Data Engineer
2024 - 2024
Komorebi Investment Pty Ltd
- Developed efficient ETL processes to ensure the AI's data was always fresh and relevant, enhancing its ability to provide accurate and contextually relevant information.
- Built AI assistants using retrieval-augmented generation (RAG) techniques, integrating large language models like GPT-4-turbo with vector databases such as ChromaDB and Pinecone.
- Worked with a South African company to develop AI assistants using retrieval-augmented generation (RAG) techniques, integrating large language models like GPT-4-Turbo with vector databases such as ChromaDB and Pinecone.
Technologies: Data Science, ETL, Databases, ETL Tools, Data Cleaning, Data Pipelines, Llama
Data Scientist | ML Engineer
2023 - 2024
ContractPod Technologies Limited
- Developed and fine-tuned advanced models for tasks such as document classification, summarization, information discovery, and risk analysis using models like GPT-4, Llama 3, Gemini 1.5, and Mistral.
- Fine-tuned open-source models like Llama 3 to match the performance of proprietary models such as GPT-4, addressing privacy concerns and ensuring data security without compromising performance.
- Optimized inference costs by improving prompt efficiency and reducing LLM calls, enhancing both the privacy and performance of AI models and resulting in substantial cost savings.
Technologies: Natural Language Processing (NLP), Data Science, LSTM, Machine Learning, TensorFlow, BERT, Python, OpenAI, Llama 2, Reinforcement Learning, Large Language Models (LLMs), Language Models, ChatGPT, Falcon, PEFT, Llama, Hugging Face
AI Engineer | Machine Learning Engineer
2017 - 2023
Mahisoft
- Engineered and optimized named entity recognition model using NLTK, scikit-learn, and SpaCy. This enhanced solution enhanced simplicity, speed, and reliability, successfully replacing the previous production model.
- Designed and developed a sophisticated prototype utilizing computer vision and deep learning techniques, seamlessly integrated into the accompanying software suite of cameras manufactured by a renowned Japanese electronics company.
- Automated the integration of our systems with various third-party machine learning solutions, including Google's AutoML. Simplified the process, enhancing efficiency and enabling seamless adoption of advanced technologies.
- Developed a data ingestion pipeline that breaks down legal documents stored in diverse formats, generating multiple artifacts to facilitate future and further experimentation.
- Developed a Java and Spring Boot-based RESTful framework to train NLP models for tasks like text classification, clustering, and named entity recognition. This sped up model development, resulting in improved performance and scalability.
Technologies: Natural Language Toolkit (NLTK), Computer Vision, Generative Pre-trained Transformers (GPT), Natural Language Processing (NLP), AutoML, Pandas, Keras, NumPy, Scikit-learn, TensorFlow, Java, Spring Boot, Python, Machine Learning, PySpark, Generative Pre-trained Transformer 3 (GPT-3), Artificial Intelligence (AI), OpenAI, ChatGPT, OpenAI GPT-4 API, OpenAI GPT-3 API, SQL, Language Models, Neural Networks, Azure, Data Pipelines, ETL, Databases, Hugging Face
Machine Learning Engineer
2019 - 2022
Blue Orange Digital
- Implemented the data model of an NLP-powered application to automatically tag topics of interest in pharmaceutical documents, supported by a combination of key terms and machine learning, surpassing the accuracy of human reviewers by 10%.
- Implemented a forecasting model using SARIMAX and traditional machine learning to help the sales team of an important pharmaceutical company project quarterly earnings.
- Created a computer vision solution to correct poor-resolution PDF scans, which resulted in a higher-quality text extraction through OCR.
- Maintained a Spark-powered data aggregation platform to consolidate information about high-profile brokers used by one of the biggest investment firms in the United States.
Technologies: Scikit-learn, TensorFlow, Natural Language Toolkit (NLTK), OpenCV, Data Science, Computer Vision, Generative Pre-trained Transformers (GPT), Natural Language Processing (NLP), Python, Machine Learning, PySpark, Artificial Intelligence (AI), ChatGPT, OpenAI GPT-4 API, OpenAI GPT-3 API, SQL, Language Models, Neural Networks, Data Pipelines, ETL, Databases
Senior Software Engineer
2017 - 2020
Mahisoft
- Developed a customizable password policies feature in an authentication microservice that is being used in the production environment of four major clients.
- Developed and co-designed a flexible, generic, multi-channel notifications microservice that was adopted in multiple clients' production systems.
- Worked intensely on the four important microservices of the company's flagship product, which contributed to the acquisition of several new, big, and important clients.
- Created written and multimedia documentation of multiple parts of the internal architecture of our main product, which reduced the ramp-up time of new hires by 50%.
- Provided intensive and detailed training for a new team member, which resulted in a seamless role transition in a core project for a key client.
- Imparted a series of workshops focused on the development of machine learning solutions using JVM technologies.
- Optimized the database queries of a major multi-billion dollar client that helped to reduce the runtime of a critical process from two hours to nine minutes, on average.
Technologies: Elasticsearch, Apache Kafka, MySQL, Go, Spring Boot, Docker, Java, JavaScript, SQL, ETL, Databases
Computer Vision Engineer
2019 - 2019
Freelance
- Developed a people counter algorithm that works on RGB data from cameras installed in zenithal position, which is now being used in multiple venues throughout Colombia.
- Built a height estimator based on depth sensor information which was used to prevent the access of underage people to age-restricted areas. This solution is deployed in more than ten locations in Colombia.
- Created an algorithm to extract, correct, and merge data from an RGB-D camera, which reduced the setup and familiarization time to this particular brand of cameras.
Technologies: NumPy, Scikit-learn, Keras, Python, OpenCV, Machine Learning, Databases
Back-end Engineer
2016 - 2017
Wivo
- Worked on the development of a new version of the company's main product, replacing the monolithic architecture of the application with a microservices oriented one, which reduced the maintenance time by 50%.
- Participated in the screening and interview process of dozens of potential candidates, which resulted in the hiring of four new, highly-skilled key members for the back end, front end, and infrastructure teams.
- Designed a sandbox-like architecture of microservices based on Docker and configurable templates that allowed the company to set up systems for new clients in a fraction of the time that was needed before.
- Documented the most relevant parts of the back-end ecosystem, which greatly improved the effectiveness and productiveness of the team.
Technologies: GraphQL, Python, MongoDB, Redis, PostgreSQL, Akka, Scala, Docker, Clojure, JavaScript
Senior Software Developer
2015 - 2016
Mesfix
- Migrated the previous monolithic architecture to a microservices-oriented one, which increased the scalability, response time, and overall quality of the company's product.
- Scaled and maintained the CI, test, development, and production environments.
- Led a team of three members focused on back-end development endeavors.
- Imparted a series of workshops focused on learning quickly the basics of Python, which increased the team's productivity.
Technologies: Software Engineering, Java, MongoDB, Jenkins, RabbitMQ, Scala, PostgreSQL, Python, JavaScript
Software Engineer
2014 - 2015
S4N
- Participated in the development of the flagship product of a very important client, consisting of a secure, cloud-based, vault-like storage system.
- Implemented an online, stateful onboarding form for the biggest insurance company in Colombia (Sura), which reduced the time taken by the process from one week. on average, to only three hours.
- Fostered a culture of functional programming based solutions in the company, through a series of educational, hands-on workshops.
Technologies: Software Engineering, Java, MongoDB, PostgreSQL, Akka, Scala, JavaScript
Experience
Virtual Concierge for Lifestyle Magazine
https://nande.co/Designed and implemented a sophisticated chatbot solution for a renowned lifestyle online magazine, serving as a virtual concierge to cater to user needs. The primary objective was to curate custom itineraries and provide comprehensive activity summaries for various metropolitan areas, including Chicago, based on user-supplied preferences and parameters.
This project's successful development and deployment resulted in a major increase in user engagement and satisfaction.
This project's successful development and deployment resulted in a major increase in user engagement and satisfaction.
Vehicle Detector
https://github.com/jesus-a-martinez-v/vehicle-detectionI created an advanced algorithm for car detection in video streams, capable of accurately identifying vehicles of varying sizes and distances. The algorithm leverages a combination of traditional computer vision techniques, including Histogram of Oriented Gradients, and machine learning algorithms, such as Linear SVC. The algorithm employs a sliding window approach to enhance precision and mitigate false positives and duplicate detections and incorporates a configurable heatmap based on the last N frames.
Lane Lines Finder
https://github.com/jesus-a-martinez-v/advanced-lane-linesDeveloped a robust pipeline utilizing sophisticated computer vision techniques to accurately track lane lines on road surfaces and provide essential curvature information.
The pipeline integrates cutting-edge methodologies, including perspective transform, camera calibration, edge detection, Gaussian blurring, and polynomial fitting, to achieve precise lane line detection and analysis.
The pipeline integrates cutting-edge methodologies, including perspective transform, camera calibration, edge detection, Gaussian blurring, and polynomial fitting, to achieve precise lane line detection and analysis.
Education
2008 - 2013
Bachelor's Degree in Computer Engineering
Universidad Simón Bolívar - Caracas, Venezuela
Certifications
SEPTEMBER 2018 - PRESENT
Computer Vision Nanodegree
Udacity
FEBRUARY 2018 - PRESENT
Deep Learning Specialization
Deeplearning.ai via Coursera
JANUARY 2018 - PRESENT
Machine Learning Engineer Nanodegree
Udacity
JUNE 2017 - PRESENT
Deep Learning Engineer Nanodegree
Udacity
Skills
Libraries/APIs
Keras, LSTM, Scikit-learn, NumPy, Pandas, OpenCV, Natural Language Toolkit (NLTK), TensorFlow, PyTorch, PySpark, Hugging Face Transformers
Tools
AutoML, ChatGPT, PyCharm, RabbitMQ, Jenkins, IntelliJ IDEA, Scikit-image, Amazon SageMaker, DeepSeek
Languages
Python, Java, SQL, Scala, JavaScript, Clojure, GraphQL, Go, Python 3, Falcon
Paradigms
ETL, Functional Programming, REST
Storage
Data Pipelines, Databases, MySQL, PostgreSQL, Redis, Elasticsearch, MongoDB, NoSQL
Frameworks
Spring Boot, Akka, Spark, LlamaIndex, DSPy
Platforms
Linux, Docker, Apache Kafka, MacOS, Azure
Other
Machine Learning, Software Engineering, Natural Language Processing (NLP), Data Science, Artificial Intelligence (AI), OpenAI, Neural Networks, Large Language Models (LLMs), Retrieval-augmented Generation (RAG), LangChain, Llama, Computer Vision, Deep Learning, Generative Pre-trained Transformers (GPT), Generative Pre-trained Transformer 3 (GPT-3), OpenAI GPT-4 API, OpenAI GPT-3 API, Language Models, Hugging Face, Chatbots, Generative Design, FastAPI, BERT, Llama 2, Reinforcement Learning, PEFT, ETL Tools, Data Cleaning, Machine Learning Operations (MLOps), Vector Databases, Real Estate
Collaboration That Works
How to Work with Toptal
Toptal matches you directly with global industry experts from our network in hours—not weeks or months.
1
Share your needs
Discuss your requirements and refine your scope in a call with a Toptal domain expert.
2
Choose your talent
Get a short list of expertly matched talent within 24 hours to review, interview, and choose from.
3
Start your risk-free talent trial
Work with your chosen talent on a trial basis for up to two weeks. Pay only if you decide to hire them.
Top talent is in high demand.
Start hiring