Experience

MY INFINITY VISION

• Brazil
• IT System Custom Software Development
• 1 - 100 Employee

• Senior Software Developer

  • Jan 2021 - Present

  Development of vision bases systems for the Retail Industry by leveraging machine learning. - Design of end to end system architecture for real-time systems - Design of mission critical systems - Implementation of online learning in open set machine learning problems - Implementation of training and inference of deep metric learning algorithms for real-time inference - Implementation of training and inference of deep learning based object detection algorithms for real-time inference - Implementation and optimization of real-time memory efficient feature search - Implementation and optimization of real-time multi-process pipeline in Python - Implementation of customizable and reproducible GPU based system builds via Ansible - Implementation of log obsevability and system metrics - Requirements management Tech stack: Linux, Docker, Docker-Compose, Ansible, MongoDB, NoSQL, Python, Cython, Tensorflow, PyTorch, Keras, Sklearn, TensorRT, OpenCV, Pillow, Flask, Pandas, Seaborn, NumPy, CUDA, Bash Show less



xmobots - drones

• R&D Software Developer

  • Aug 2020 - Jan 2021

  Automation of fotogrametry pipeline for monitoring of large preservation areas, illegal selective logging and dam damage detection through aerial surveys.- Computer vision based task automation- End to end photogrametry pipeline automation for large scale aerial surveys- Geographic information system (GIS) development and augmentation- RAW image processing for multispectral cameras- Sensor data format stadarization- XMP metadata manipulationTech stack: Python, C++, Java, Cython, Git ,Exiftool, Metashape, DCRaw, RawLib, Tifflib, GDAL, selenium-python, OpenCV, Metashape API (Python), PyQt5, PySide2, OpenMP (C++), Swing (Java), RTKLIB, IBGE-PPP, GPS. Show less

• Jr. R&D Software Developer

  • Jul 2019 - Jul 2020

• Trainee - R&D Software Development

  • Apr 2019 - Jul 2019



Embrapa

• Brazil
• Farming
• 700 & Above Employee

• Trainee - R&D Software Developer

  • Jan 2017 - Dec 2018

  Development of a low budget embedded real-time solution for vegetation health analysis of corn plantations using drone aerial surveys. During the project the team implemented real-time algorithms for mosaicing the pictures taken and calculating vegetation indexes based on RGB imagery. The resulting mosaic and report were displayed on a tablet on landing along with recommendations for possible interventions to improve the yield quality and quantity. All the processing was done on board in a Qualcomm Flight Development Board using Linaro Linux as the OS. This project was the result of partnership between Embrapa and Qualcomm. - Android application development - Development of embedded image processing algorithms - Implementation of machine learning inference on embedded devices - Implementation of embedded vegetation indexes and reporting - Mission Planer plugin development for drone route control and mission related metadata input Tech stack: Python, C++, Java, Cython, Git ,Exiftool, Metashape, DCRaw, RawLib, Tifflib, GDAL, selenium-python, OpenCV, Metashape API (Python), PyQt5, PySide2, OpenMP (C++), Swing (Java), RTKLIB, IBGE-PPP, GPS Show less