supportbot initial writeup

Author: rampantmonkey

resume.html

@@ -20,6 +20,8 @@ <h5>Time Series Feature Store</h5>
 <p>Data is ingested from a variety of sources - SNS, MQTT, Kinesis, and S3. The service can currently serve 2k requests per second per core in the EC2 autoscaling group with further optimizations in the works.</p>
 <h5>Temperature Compensation</h5>
 <p>Our thermostat contains many heat generating components - cellular radio, processor, and power management. These components introduce heat into the enclosure which can cause inaccuracies in the temperature reading from our on-board temperature sensor. In collaboration with our electrical engineer I designed an ongoing experiment to capture data with reference temperature sensors and internal temperature readings from each component in a variety of locations and power usage scenarios. Then I developed a pipeline on top of the time series feature store to build a model that compensates for this heat to produce a more accurate temperature reading.</p>
+<h5>Supportbot</h5>
+<p>Our engineering and support teams need to gather data from a collection of locations in order to fully understand what is happening with our devices in the field. While we continue to extract the common questions and build those workflows into our existing product dashboard, there are always new ad hoc inquiries. I built an agentic system and an MCP server to be a support research assistant. Now we are able to ask free-form questions about the health or performance of our devices, request recommendations for improvements to thermostat settings, and diagnose ongoing issues with less manual effort and less specialized knowledge about HVAC, IoT, and thermodynamics.</p>
 
 <h3>WireWheel Inc. - Lead Architect - 11/2020-02/2022</h3>
 <blockquote> WireWheel offers products for achieving compliance with privacy laws while preparing your organization to respond to upcoming privacy standards.  </blockquote>