Michael Pesavento Email & Phone Number

Berkeley, CA, US

Consulting work available in the areas of machine learning, computer vision, natural language processing, signal processing, biotechnology, embedded devices and sensors, and more.

Provided technical leadership for software system design and execution, synchronizing sound and light in real time in a one-of-a-kind suspended spherical auditorium with spatialized sound.

Seattle, WA, US

Provided technical leadership and system design for the Accessibility Engineering Team, while focusing on innovative machine learning solutions for enhanced accessibility and user experience.

New York, NY, US

Creating Natural Language Processing models and data pipelines to process electronic medical record information

San Francisco, California, US

Technical leadership for statistical validation of autonomous vehicle system safety, reliability, and comfort

San Francisco, California, US

• Lead early analysis team to create functional data products for customers by guiding development of machine learning algorithms to enhance and automate computer vision analysis
• Co-created volumetric analysis pipeline and computer vision stack used on micrometer-resolution, teravoxel scale volumetric images generated by 3Scan’s automated microtome + microscope (Python, Java)
• Designed and created microvascular tissue identification and measurement product that identifies statistical and structural differences in vascular networks between and within whole organ image volumes
• Created image data products for customer contracts, created database system to track data product location on AWS cloud storage, updated browser-based 3D pyramidal tile viewer to view generated data products
• Worked closely with other divisions to monitor image quality control and identify key performance metrics, increasing image throughput and quality for customer project data.

US

• Developed algorithms for 3-axis accelerometer motion processing (step count, distance, calories) for use in ARM-based embedded systems (C)
• Managed project team that created cloud-based web applications for displaying user fitness data & analysis
• Created biosignal processing algorithm that estimates stress level based on ECG data from wearable biosensors
• Tested and improved the product experience of EEG algorithms by reducing calibration time and improving comparison of mental state across time
• Designed and performed statistical validation on 3rd party libraries for in-house motion processing and ECG used in product algorithms
• Developed motion data collection framework using the Android SDK and Bluetooth

• Developed audio waveform processing algorithms and models for sound analysis and segmentation (Python)
• Researched and improved techniques in speech/noise segmentation in push-to-talk radio, telephone, and microphone environments
• Improved accuracy of sound segmentation statistical models by implementing novel features

La Jolla, CA, US

• Researched neuronal responses in primary somatosensory cortex to in-vivo rodent whisker deflections via two-photon laser microscopy, activity-dependent fluorescent dyes, electrophysiological recordings, and behavioral.
• Developed software for capture and analysis of intrinsic optical signal imaging; developed system hardware and interface and calibration that ensured data quality
• Developed analysis software for rapid off-line statistical analysis of video & analog data streams
• Augmented computer vision boosting algorithm for automatic cell type discrimination and segmentation

Rochester, NY, US

• Created real-time C++ simulations of conductance-based computational models of excitatory and inhibitory neuron populations of rodent primary somatosensory cortex.
• Interfaced real-time simulations with simultaneous neurophysiological recordings in living neural tissue.
• Utilized non-linear dynamic system analysis to examine relationship between simulation and biological circuit.
• First author in peer-reviewed publications; presented results at 8 international neuroscience conferences.
• Trained and supervised undergraduate staff in neurophysiological preparations, recordings, and data analysis

Los Angeles, CA, US

* Developed MATLAB software for use in examining visual perception of synchrony in human subjects; developed computational models of unintentional errors in eye movement.* First author of a peer-reviewed publication; presented results at 2 international neuroscience conferences.

Doctor Of Philosophy (Phd), Neurobiology & Anatomy

Master Of Science (Ms), Neurobiology And Anatomy

Master Of Science (Ms), Biomedical Engineering

B.S., Cognitive Science/Neuroscience