Allay Desai
- +1-716-868-5405
- allaydesai@gmail.com
- www.allaydesai.com
- Baltimore, MD
A self-motivated software engineering professional experienced leading technical project teams in delivering robotics and automation system programs across a diverse spectrum of industries. An innovative problem-solver skilled in the full lifecycle of deploying artificial intelligence, deep learning, and machine learning initiatives. Leverages strategy and execution competencies to formulate plans and drive bottom-line results. Operates with an entrepreneurial spirit, a Master’s Degree in Electrical Engineering along with numerous industry certifications.
Bilingual: English | Hindi | Arabic
Work Experiences
Senior Software Engineer
As a Sr. Software Engineer my role is to work with a team of software developers for industrial automation applications in variety of domains. This includes leadership role in overseeing junior developers and guiding development cycle from conception to completion.
- Design and develop PLC software for industrial applications.
- Create WPF applications for windows in C# using data binding and MVVM framework for system user interface.
- Use machine learning algorithms such as Neural Nets, Naïve Bayes, Random Forest and Support Vector Machines for smart machine controls.
- Oversee data collection, cleaning and transformation for deep learning models along with feature selection.
- Develop and deploy deep learning models for various image classification and localization applications.
- Implement Behavioral Cloning to model system control based on data collected in ideal situation.
- Implement computer vision algorithms for feature extraction, object detection and data augmentation.
- Program Arduino microcontroller and raspberry pi in C++ to implement control system components.
- Develop flow charts and UML diagrams to represent software design.
- Write unit tests in NUnit framework for Test Driven Approach in software development within visual studio.
- Work with RGB GigE Vision cameras and Intel RealSense 3D camera to build image acquisition pipeline.
Electrical Design Engineer
As an Electrical Design Engineer my role was to design and develop plc logic for industrial systems as well design panels and electrical schematics. This included prototyping, building and testing in-house followed by commissioning system in field.
- Design control logic for complex industrial and commercial automation systems in material handling domain.
- Program a Programmable Logic Controller (PLC) to achieve desired machine functionality.
- Design panel and safety circuits in AutoCad Electrical for industrial machines to comply with safety standards.
- Electrical wiring of panels, sensors and actuators as well as assist building of machines in the workshop.
- Review mechanical designs in Autodesk Inventor and participate in system design brainstorming sessions.
MEP/Energy Services Intern
Team member of highly skilled professionals offering consulting engineering for MEP/Energy services. Performed research and assisted in preparation of specifications, bid documents, letters and reports based on site visits for various scopes of work. Created an automated system for capturing data from a building energy monitoring system using Raspberry Pi mini-computers along with a database of ASHRAE Level I, II and III energy audits for energy benchmarking and opportunity assessment.
Rotational Trainee Engineer
As a trainee engineer, I performed data analysis of past 3 years of HDPE and PET bottles manufacturing data with the goal of implementing statistical process control using the WinSPC software. Further, I observed and studied the process control of extrusion blow molding and injection blow molding while conducting a case study for feasibility of color sensor in current manufacturing line to detect color defect.
Projects
Self-Driving Car Projects, Vehicle Detection, Tracking & Planning
- Lane detection pipeline built by applying computer vision techniques.
- Detection and Classification of traffic signs and traffic lights using Single-Shot Detection (SSD) model.
- Utilizing both Extended (EKF) and Unscented (UKF) Kalman Filter to estimate the state of a moving object of interest with noisy lidar and radar measurements.
- Perform sparse localization using a 2-dimensional particle filter and combine it with a real map to localize a vehicle.
- Design PID and MPC Controllers to drive a car around a simulated track.
- Implement a path planning module which generates smooth safe trajectories for a car to follow on a highway.
- Perform scene understanding using semantic segmentation to label the pixels of an image in categories such as road, car, pedestrian, sign.
- Utilize Robot Operating System (ROS) as an environment to create perception, planning and control module of Carla, Udacity’s Self Driving Car.
Deep Learning Projects, Classification and Natural Language Processing
- Generate faces: Built a pair of Multi-Layer Neural Networks and make them compete against each other in order to generate realistic faces.
- Dog Breed Classifier, CNN: Designed and trained a convolutional neural network to analyze images of dogs and correctly identify their breeds.
- Vehicle Lane Detection, Computer Vision & DL: Wrote a software pipeline to identify the lane boundaries in a video from a front-facing camera on a car.
- Traffic Sign Classifier, CNN: Developed a deep learning model using convolutional neural networks to classify traffic signs.
Data Analytics
- Using numpy, pandas and matplotlib to investigate data associated with imdb movie reviews from the year 1960 to 2015.
- Wrangle WeRateDogs Twitter data gathered from multiple sources to create interesting and trustworthy analyses and visualizations.
- Analyze A/B test run by an e-commerce website to understand effectiveness of a new web page whose goal is to try and increase the number of users who "convert,"
- Perform univariate, bivariate and multivariate analysis on factors affecting borrower’s APR interest rate by analyzing Prosper loan data and creating a short presentation.
Computer Science and Robotics Projects
- Programmed AmigoBot by MobileRobots to wander in an unknown environment using inbuilt sonar sensors for detection of obstacles, enhanced by integrating a sound sensor and LED’s for feedback.
- Experience with other Robots such as Hexapod, Four Wheel Differential Drive, AL5B Robotic ARM, Picasso.
- Chip Design Project (VLSI), utilized Cadence Virtuoso in designing the layout of a Vending machine chip.
- Programming a game (Donkey Kong) from scratch using ARM Assembly language on an ARM7 micro-processor.
- Building Energy Monitoring System deployed on a raspberry pi which hosts a web server that gathers and analyzes data from ZigBee sensors