Machine Learning, Data Scientist

Enoch Sarku

Email: **************@*****.*** Tel: +1-253-***-****)

SUMMARY

Experienced professional efficient with statistical analysis, Machine Learning and Pattern Recognition, R, Python, SQL, Teradata Mining, MLOps, Docker, deep learning, C++ Programming, Computational Modelling in MATLAB, Linux, Bash Shell scripting, Robotic Operating System (ROS), Azure, AWS Lambda, AWS EC2 instance, AWS Sagemaker (Studio, Canvas), GitHub, Jira, Agile, Git, Pytorch, Google Cloud Platform, Actions on Google, TABLEAU and Project management.

PROJECTS and SPECIALTIES

Designed interactive Ontime and flight delay assessment using SAP business objects and queried Sam’s Club data warehouse using SQL and Teradata.

Employed SAP in Academic Research to analyze agritourism data

Queried food bank data using SQL in R language to help predict demand for food for inventory stocking

Used AWS Lambda function to build a Json data extraction pipeline through Textract Analyze and make API calls.

Setup domain and subdomain routing in AWS Route53 for a startup

Deployed a Frontend (React) and Backend (Spring Boot Application): Setup security groups, VPC and successfully provisioned EC2 instance to deploy React website and Spring boot-java backend using ssh to transfer compiled projects for a startup and implemented a CI/CD pipeline using Jenkins.

Used Langchain/Llamaindex and OpenAI model API to design question answering pipeline for 1400-page USCIS policy document.

Sentiment Analysis of Algorithmic Bias in Lending: Performed sentiment analysis on scraped Facebook posts for perceptions of algorithmic bias in lending to small businesses using tools like pandas, transformers - pipeline package and nltk- sentiment intensity analyzer.

Classifying visas: Utilized large language models to classify visas based on a few sets of inputs in Langchain and LlamaIndex with a vector database (vectorstore and chromadb)

Used K-Means, KNN and Autoencoders with TensorFlow in Google Collab and python SciPy to detect fraud for computational statistics class project.

Used Takt time, forecasting models, time study, queuing theory, statistical process control and Break-Even analysis to determine volume and pricing for a start-up.

Used Fick’s Law of Motion and finite difference to model Epidemic dispersion in 2-Dimensional space in MATLAB.

Used rvest and tidyverse to mine and transform web data and apply predictive modelling in R on census income data.

Used SciPy, Rasterio, & PILLOW libraries in python and Jupyter Notebook to detect, classify and cluster buildings in satellite image.

Developed a code in a team to determine the atomic distance between molecules to derive a potential collagen nanofiber variant.

Other skillsets include setting up Azure VMs and using versioning (git) to write, test and privately deploy code for autonomous vehicle models using streamlit.

EDUCATION

North Carolina Agricultural & Technical State University, NC, USA

PhD Computational Science and Engineering- 2020 to Current

North Carolina Agricultural & Technical State University, NC, USA

M.S. Technology Management, Jan 2019 – Dec 2019: Switched to PhD

North Carolina Agricultural & Technical State University, NC, USA

M.S. Agribusiness & Food Industry Management, Aug 2015 - May 2017

University of Ghana, Ghana

B.S. Agricultural Economics, Aug 2008 - May 2012

PROFESSIONAL EXPERIENCE

LiDAR Consultant- (Nissan Autonomous Vehicle Research) - Kelly Services Dec 2022 – Sep 2023

- Researched and developed evaluation metrics and wrote python code to evaluate LiDAR tracking model performance

- Created Azure VM’s for cloud collaboration and Story boards using Jira to streamline solutions

- Used Docker to deploy evaluation and visualization tools

Perception Research Intern - General Motors Jun 2022 - Sep 2022

-Applied Image processing techniques to extract and process depth information from a Time-of-Flight 3D camera sensor in MATLAB and python

- Applied regression techniques to validate spatial uniformity of point cloud data from Time-of-Flight (ToF) camera sensor

- Estimated temporal error using pixel wise comparison and characterized Time of Flight (ToF) camera sensor for performance

-Estimated plane from camera point cloud data using plane regression techniques and developed 3D Scene Reconstruction for Autonomous Vehicle Applications

- Developed partial end to end software stack for Depth Camera Based Sensor

Student Engineering Lead & Technical Project Manager -SAE-GM Autodrive Challenge (NCA&T A3 Team)

SAE International Jan 2022 - Jun 2022

- Applied Image processing methods to build lane detection models from a gigabit camera sensor

- Led a team to develop perception, design and Electrical system for General Motors Chevy Bolt vehicle and transform chevy bolt EUV prototype into an autonomous vehicle

- Leading development of software solutions for autonomous vehicle navigation using docker, python, and MATLAB ROS

- Managing procurement and negotiations with third party vendors

-Using project management tools like Jira, agile, WBS to schedule team effectiveness

-Managing and teaching software tools such as GitHub, Azure Dev-ops, Build Tools, image processing techniques, computer vision

and autonomous driving architectures

Perception Lead- NCA&T State University A3 Team

SAE International Oct 2021 - Jun 2022

- Led a team to design camera, radar and LiDAR perception algorithms for objects classification, traffic signs classification and localization of ego vehicle

- Developed Object distance detection from point clouds using pcl library in python

-Determining sensor and compute requirements and Procurement of Sensors from suppliers through

quotes, POs and order negotiations by phone and email

Graduate Research Assistant & GM Auto drive Challenge Team Member

North Carolina Agricultural & Technical State University 08/01/ 2019-Oct 2021

Detecting 10 traffic sign classes using Transfer learning with Pytorch Resnet models, CNN and TensorFlow

Successfully developed Motion detection, low level firmware’s and an object detection algorithm in 3D material printer droplets using OpenCV and morphological transformations

Developing algorithm for droplet size detection and droplet regulation automation in 3D material printer using OpenCV, moment invariants and morphological transformations

Implementing Lidar segmentation for Autonomous car navigation in ROS and C++

Building thermal camera temperature sensor automation using machine learning and motion detection in OpenCV and c++

Developing and implementing localization and path planning for autonomous vehicle in ROS

Data Analyst

Address: 1601 E Market St, Greensboro, NC 27411, USA

North Carolina Agricultural & Technical State University Cooperative Extension 08/14/2015-12/08/2019

Leveraged TABLEAU software and Excel to identify and classify small farms by distribution across specific income classes in all 100 NC counties.

Created interactive data maps to aid new staffing & program decision, performed thorough data analysis, interpretation and presentation leading to 3 new staff hired and 2 programs created.

Collaborated and won Value -Added research funding proposal in a team of six.

Presented economics research poster – Challenges of Adopting Agri-tourism in NC – in Texas during graduate studies leveraging Logistic Regression in Excel and SPSS

PUBLICATIONS

Reducing Data Costs-Transfer Learning Based Traffic Sign Classification Approach

Non-invasive Low-Cost Fever Detection Systems

Predicting heart Disease using Machine Learning

Optimal Features for Cross Subject Classification of Imagined Left and Right Fist Movements using EEG Signals

Factors influencing adoption of Agri-tourism among smallholder farmers in NC.

Link to publications: https://scholar.google.com/citations?user=s-qOY9wAAAAJ&hl=en

Certifications:

AWS Cloud Practitioner

AWS Sales

Tableau

Six SIgma

REFERENCES

Available upon request

Contact this candidate