Gant Kulnirundorn Email & Phone Number

San Francisco, CA, US

• Solving the challenge of regional expansion of Databricks offerings in multi-cloud environments. Built a framework to capture and organize complex deployment steps and dependencies across the entire company.
• Developed a unified system to monitor, automatically remediate, and send real-time alerts, ensuring Cloud resource usage remains within set limits for all Cloud Providers. Ensured availability of key features, such as.

Mountain View, CA, US

• Tech Lead Manager of an infra team within Sensor Algorithms group. The group deliver sensor algorithms behind features like fall detection, gesture recognition, step counter, and distance tracking found in Google.
• Led the team to build reusable tools and platforms, including sensor data collection apps, data labeling tools, and offline evaluation frameworks, accelerating algorithm development cycles. Widely adopted by engineers.
• Prototyped, benchmarked, and proposed the group to adopt TensorFlow Lite Micro (TFLM) as a standard way to run on-device inference of ML-based algorithms.

Mountain View, CA, US

• Built Speech Annotation Platform, a core system enabling Speech researchers and engineers to obtain high quality training and testing datasets for building Google’s voice recognition features (such as "Ok Google").
• Developed and implemented a strategy to mitigate quality issues in the audio labeling system often caused by user errors and malicious actors who abused the system.Speech Data Infrastructure Team

Mountain View, CA, US

• Employed by Kelly Services, supporting Google.
• Built web applications and big data pipelines (MapReduce and Flume) for Speech data preparation, cleaning, and annotations. Worked with a distributed team of linguists in multiple countries.

Toronto, ON, CA

• Expertise in High-Performance Market Data APIs: Resolved complex performance issues, including concurrency issues, memory leaks, and process crashes for Thomson Reuters customers. Used low-level debugging tools such as.
• Built a tool that reassembles raw TCP/IP traffic of company's proprietary RSSL protocol into human-readable format. Enabled the teams to solve problems using clients' network capture files.
• Built a log file analyzer that could handle large log files efficiently. The tool helped a team of 10 consultants identify root causes for problems faster.

Armonk, New York, NY, US

• Developed a proof-of-concept density calculation service for CMOS chip. Web application using Struts framework, Java (J2EE).