Matthew Tognotti

Software Engineer

I recently graduated from Santa Clara University with a degree in Electrical and Computer Engineering. For two years, I contributed to the Human-Machine Interaction & Innovation Robotics Lab at SCU as a software engineer and research assistant.

During my time there, I focused on developing a speech-interfaced mobile manipulator. You can see the robot in action in the videos below and find our publications here.

Work Experience

Human-Machine Interaction and Innovation Robotics Lab

Sept. 2023 - Sept. 2024

Software Engineer & Researcher

Led a project funded by the National Science Foundation's Research Experience for Undergraduates to develop a robot learning system for a mobile manipulator to learn and execute user-taught pick-and-place tasks.

Improved system performance by 67% through iterative design optimization and user feedback integration, delivering reliable automation that met real-world operational needs.
Authored and presented a conference publication at IEEE RO-MAN, cited in [1].
Videos of the robot and the learning system are available here

June, 2023 - Sept., 2023

Kuehler Award Researcher

Collaborated with PhD students to develop a speech-to-action robotic system, enabling a mobile manipulator to identify, pick, and place items based on speech commands using Robot Operating System (ROS), Google Speech-to-Text, GPT-Neo, and Intel RealSense RGBD Cameras.

Earned the Kuehler Research Award for novel contributions to robotic human interaction and co-authored three peer-reviewed publications listed below.
Designed and conducted user studies to evaluate system effectiveness and usability, incorporating stakeholder feedback to optimize solution performance.

Jan, 2023 - June, 2023

Research Assistant

Developed an API in Python and ROS to create seamless integrations between hardware and software systems.
Created kinesthetic teaching capabilities using MoveIt and teleoperation for a manipulator.
Built a virtual lab in RViz with simulated obstacles and 3D meshes to test autonomous path planning.

Education

Santa Clara University

B.S., Electrical & Computer Engineering

September 2024

GPA: 3.65/4.0; Cum Laude

Coursework: Robotics, Data Structures & Algorithms, Real-time Embedded Systems, Computer Architecture, OS

Publications

International Conference on Robotic Computing (IRC)

[1] Tognotti M. and Kyrarini M., 2024. Are you Ready? an Intelligent Robotic Assistant for Instrumental Activities of Daily Living. In 2024 Eighth IEEE International Conference on Robotic Computing (IRC), pp. 249-252, doi: 10.1109/IRC63610.2024.00038
International Conference on Automation, Robotics and Applications (ICARA)

[2] Kodur K., Zand M., Tognotti M. and Kyrarini M., 2024. Translucent Object Grasping Using Robot Vision. In 10th International Conference on Automation, Robotics and Applications (ICARA), pp. 107-111, IEEE. doi: 10.1109/ICARA60736.2024.10552988
A User Study

[3] Kodur K., Zand M., Tognotti M., Jauregui C. and Kyrarini M., 2023. Structured and Unstructured Speech2Action Frameworks for Human-Robot Collaboration: A User Study. doi: 10.36227/techrxiv.24022452

Software Projects & Entrepreneurship

Matt's Matcha - E-commerce and Retail Brand

Co-founder/Generalist

Launched a direct-to-consumer brand from scratch, handling everything from legal setup and supply chain management to branding, packaging, customer engagement, and retail operations.

Designed and customized a Shopify storefront, led marketing efforts (SEO, outreach, social media), and created all product visuals and content.
Operated with startup-level speed and adaptability – self-taught across functions, iterated quickly, and executed end-to-end with limited resources.
View our site here

Automated Stock Checker Bot Deployed to AWS

Built a Python-based bot using Selenium that automates login flows, bypasses reCAPTCHA, and continuously monitors product stock – sending real-time Telegram alerts with product details (price, size, link).
Designed with clean OOP architecture (e.g., StockChecker , TelegramBot classes) for modularity and extensibility; currently deploying to AWS (EC2 + CloudWatch) for 24/7 uptime and production monitoring.
View on GitHub

LLM Tool-Calling Agent with Retrieval-Augmented Generation

Built a GPT-4-powered agent leveraging OpenAI function calling to execute real-world tasks (e.g., sending emails), designed as a foundation for multi-tool LLM systems.
Currently extending to support Retrieval-Augmented Generation (RAG) and model context protocol (MCP), enabling complex task delegation and context-aware reasoning.