unspun is a venture-backed robotics and digital apparel company building custom jeans for each consumer, on demand.

Our mission is to reduce global carbon emissions by at least 1% through automated, localized, and intentional manufacturing. Our revolutionary technology currently fuels our own custom denim label unspun, a software API platform powering digital customization for multiple large apparel brands, and a National Science Foundation-backed hardware team working on the future of additive and automated apparel production.

Food isn’t trash. I’ve worked on creating the first flagship product from the ground up, the Mill kitchen bin.

* DRI on cross-functional workstreams for core features; mass sensing and dry-grind. Utilized lab

experiments and field-trial data to iterate quickly over designs and free parameters. Python, C, & SQL

* Designed, developed, and deployed sensing, signal processing, and control algorithms. Python & C.

* Set up monitoring to validate the correct operation of deployed algorithms in the field. Python & SQL.

* Used field data to identify and flag bugs; developed lab experiments and created proxy measures to

monitor the health of sub-systems when no reference signal is available. Python & SQL.

* Provided mentorship to teammates in mathematical computation and in preparing proposals and presentations.

* Supported product team to make data-driven decisions and interpret statistics.

Skylo is deploying the world’s most affordable and ubiquitous satellite-based IoT network that can connect any machine or sensor. My responsibilities as an Algorithm Engineer within the Advanced Networks and Algorithms Team include designing and implementing algorithms, often on resource constrained embedded platforms, to address challenges across the organization and providing specialist support to product and development teams.

* Designed and implemented an adaptive concave hull algorithm to calculate the covered area by user equipment; ~10x reduction of error against baseline over a representative test set.

* Designed and implemented an online algorithm to run on a resource constrained embedded platform to improve GNSS performance; ~4x average improvement in CEP (Circular Error Probable) over real life tests.

* Implemented an algorithm for compensating doppler shift resulting from orbital mechanics of geostationary satellites; ~7x average, ~3x worst case margin to maximum tolerable error.

* Designed an implemented a running odometer for vehicle mounted user equipment on a resource constrained embedded platform.

* Designed and implemented a post-processing procedure for GNSS data to improve user experience.

* Designed an encryption scheme for multicast messaging.

* Ownership of 3 modules related to core network, application, and embedded firmware.

* Supported international Development and Product teams with specialist expertise.

Research:

* Collaborative Trajectory Control for Aerial Networks: Decentralized multi-agent trajectory planning of UAV based access points to implement a flexible aerial Radio Access Network (RAN) that can be rapidly deployed in previously unknown environments, utilizing Network Science and Reinforcement Learning.

* SANCHO: Data informed model examining staffing strategies to handle healthcare worker absenteeism while accounting for the structured contact-network heterogeneity, with a case study of a COVID-19 Hospital in Central Florida using real data. Simulations & Analysis in Python.

* Social Bots on Twitter: Studying the implications for the significant existence of social bots on Twitter, with a focus on bot detection and data science approaches. Conducting a validation study on publicly available bot detection tools and developing an open-sourced data annotation tool. Python & JavaScript.

* Study on Social Integration of Refugees in Turkey: Using an unprocessed 1.95GB Call Detail Record (CDR) dataset from 50k users, I have constructed mobility networks based on a variable high-order network topology, and identified distinct markers for spatial segregation of refugees. Analysis in Python.

Teaching: Assisted Signals & Systems and Deep Learning classes, handling all the computational sections, assignments, and projects; often for audiences with limited to no prior programming experience.

Worked in development, deployment, and test of Health-Care IoT and Real-Time Locating (RTLS) solutions.

* Conducted R&D for a locating (RTLS) solution that is currently deployed in multiple hospitals. Worked on virtually every aspect from concept to deployment, with a focus on research and hardware development.

* Shipped 6 mass-produced RF devices: Designed, troubleshot, and validated all the circuitry end-to-end from concept to delivery for low-power wearable RF devices, sensors, and fixed RFID gateways. This project won The Best Health-Care RFID Implementation from RFID Journal Awards 2018.

* Worked together with respective coworkers to oversee mass production, develop test tools for production, improve enclosure and UX design, and develop in-house quick prototyping options.

* Designed proof-of-concept devices with short turnaround times for quick prototyping regularly.

* Done schematic capture, PCB design, and layout for circuits with RF, analog, and digital components on rigid and flexible substrates, following DFM and EMC principles.

* Experienced in RF instrumentation and use of lab equipment such as spectrum analyzers, VNAs, signal generators, oscilloscopes, logic analyzers, etc.

* Led EM and regulatory compliance efforts. Held internal lectures and trained interns on EMC concepts.

* Wrote embedded software in C for multiple MCU platforms, both on bare metal and with an RTOS.

* Developed, conducted, and documented assessment and validation tests for different RFID technologies.