Romina Abachi Email & Phone Number

Montreal, QC, CA

Toronto, Ontario, Canada

Toronto, Ontario, Canada

- Assisted in teaching a class of 6-10 industry participants the foundations of reinforcement learning, created assignments and conducted reading group sessions- Working individually with each company’s group to guide the design of a simulation for their project and implement it in Python, and to come up with possible solutions- Projects were: allocation.

- Studied literature on verifiable robustness bounds to adversarial attacks, especially those using convex relaxations and duality theory to obtain trainable robust networks with guarantees.- Improved on the state-of-the-art robustness bounds based on convex relaxations by ~8% by replacing the activation function in the dual network by a dual version of.

Greater San Diego Area

• Designed and programmed a tool in VBA for automating the calculation of trace width on a PCB.
• Performed simulations to check the resistance of traces on a PCB to ensure losses did not exceed a pre-determined margin.
• Designed a circuit to break-out GPIO connections for testing from a development platform based on the next generation of modems. This included spec'ing for parts, checking layout and stackup, and ensuring mechanical.
• Designed a test vehicle board for testing chips with very fine-pith ball grid arrays. This included test vehicle included daisy chains for three different chips on different pad geometries, as determined upon design.

• Designed Network Composite Maps using Microstation based on recent development of electrical and civil assets
• Wrote Scope documents to plan and organize asset renewal, network vault rebuilds, and conversion of legacy infrastructure
• Analyzed capacity of feeders and transformers for new customer connections and made relevant recommendations to the engineers
• Forecasting of load and planning for new future infrastructure on Eglinton St. in anticipation of the new Eglinton Crosstown LRT.

• Developed and tested code for CAN communication and data transmission amongst several microcontrollers
• Programmed different types of TI microcontrollers to perform tasks related to DC/DC conversion in mobile solar application
• Designed schematic for a supervisory board to be used in real-life application
• Devised a test platform and designed the schematic and PCB layout for current-voltage characterization of novel Gallium Nitride (GaN) HEMT transistors
• Designed complete PCB layout for two-layer and four-layer boards
• Automated testing of GaN transistors using LabView

• Currently designing a carbon monoxide and natural gas detector in a team of five
• Spec'd sensors, microcontrollers and other components to suit our application
• Designed schematic and PCB layout to perform sensing and processing in one board rather than across several off-the-shelf boards
• Currently developing code to interface a TI microcontroller with Wifi to transfer data to a remote server

Singapore

Designed, implemented, and tested a novel method of biomarker detection based on measuring current changes across a nanoporous membrane in an electrolyte solution

• Independently designed and tested a system for measurement of emissivity of thin metal surfaces as a function of surface oxidation using contactless IR spectroscopy methods
• Supervised and mentored three high school interns, developing leadership skills

Master Of Science - Ms, Computer Science

Master Of Science - Ms, Reinforcement Learning

- Supervisors: Amir-masssoud Farahmand and Brendan Frey - Thesis: “Policy-Aware Model-based Reinforcement Learning”

Bachelor’S Degree, Electrical, Electronics And Communications Engineering, 3.92

Activities and Societies: Toronto Solar Decathlon, IEEE Design Team, The Cannon school paper, Skulebook, Youth Sarcoma Initiative