Posts by Tags

FFT

LlamaIndex

OpenAI

Perplexity

Python

agents

Trying to Get Signals Agent to Work

11 minute read

Published:

For the past few weeks, I’ve been building a project called Signals Agent, a tool designed to analyze environmental audio using FFTs and other signal processing techniques, powered by an LLM-based reasoning loop. The goal was simple in theory: create an agent that could make sense of ambient sounds by performing multiscale spectral analysis, identify interesting frequency events, and describe likely real-world sources. As with any real-world agentic system, the reality has been much more complex. This blog post documents my recent efforts to improve the accuracy and reliability of the agent, especially in distinguishing between sound sources with overlapping spectral content.

audio

Trying to Get Signals Agent to Work

11 minute read

Published:

For the past few weeks, I’ve been building a project called Signals Agent, a tool designed to analyze environmental audio using FFTs and other signal processing techniques, powered by an LLM-based reasoning loop. The goal was simple in theory: create an agent that could make sense of ambient sounds by performing multiscale spectral analysis, identify interesting frequency events, and describe likely real-world sources. As with any real-world agentic system, the reality has been much more complex. This blog post documents my recent efforts to improve the accuracy and reliability of the agent, especially in distinguishing between sound sources with overlapping spectral content.

fa

fpga

TDR FPGA Progress

4 minute read

Published:

Simulating Time Domain Reflectometry (TDR) on an FPGA

2025 Summer Project Ideas

2 minute read

Published:

The past semester, I have been learning system verilog and applying it for use during labs on FPGAs. I have learned to love this process, and I want to find more ways to apply and expand on these skills. In addition, I have found interest in semiconductor fabrication tools as well as machine learning. For these reasons, I want my summer project(s) centered around FPGA, ML, and the SC industry. Outlined below are three ideas I have been playing around with.

gaussian splitting

2025 Summer Project Ideas

2 minute read

Published:

The past semester, I have been learning system verilog and applying it for use during labs on FPGAs. I have learned to love this process, and I want to find more ways to apply and expand on these skills. In addition, I have found interest in semiconductor fabrication tools as well as machine learning. For these reasons, I want my summer project(s) centered around FPGA, ML, and the SC industry. Outlined below are three ideas I have been playing around with.

ideas

2025 Summer Project Ideas

2 minute read

Published:

The past semester, I have been learning system verilog and applying it for use during labs on FPGAs. I have learned to love this process, and I want to find more ways to apply and expand on these skills. In addition, I have found interest in semiconductor fabrication tools as well as machine learning. For these reasons, I want my summer project(s) centered around FPGA, ML, and the SC industry. Outlined below are three ideas I have been playing around with.

ml

TDR FPGA Progress

4 minute read

Published:

Simulating Time Domain Reflectometry (TDR) on an FPGA

2025 Summer Project Ideas

2 minute read

Published:

The past semester, I have been learning system verilog and applying it for use during labs on FPGAs. I have learned to love this process, and I want to find more ways to apply and expand on these skills. In addition, I have found interest in semiconductor fabrication tools as well as machine learning. For these reasons, I want my summer project(s) centered around FPGA, ML, and the SC industry. Outlined below are three ideas I have been playing around with.

numpy

signal-processing

Trying to Get Signals Agent to Work

11 minute read

Published:

For the past few weeks, I’ve been building a project called Signals Agent, a tool designed to analyze environmental audio using FFTs and other signal processing techniques, powered by an LLM-based reasoning loop. The goal was simple in theory: create an agent that could make sense of ambient sounds by performing multiscale spectral analysis, identify interesting frequency events, and describe likely real-world sources. As with any real-world agentic system, the reality has been much more complex. This blog post documents my recent efforts to improve the accuracy and reliability of the agent, especially in distinguishing between sound sources with overlapping spectral content.

signals

spectrogram

summer

2025 Summer Project Ideas

2 minute read

Published:

The past semester, I have been learning system verilog and applying it for use during labs on FPGAs. I have learned to love this process, and I want to find more ways to apply and expand on these skills. In addition, I have found interest in semiconductor fabrication tools as well as machine learning. For these reasons, I want my summer project(s) centered around FPGA, ML, and the SC industry. Outlined below are three ideas I have been playing around with.

systemverilog

tdr

TDR FPGA Progress

4 minute read

Published:

Simulating Time Domain Reflectometry (TDR) on an FPGA

2025 Summer Project Ideas

2 minute read

Published:

The past semester, I have been learning system verilog and applying it for use during labs on FPGAs. I have learned to love this process, and I want to find more ways to apply and expand on these skills. In addition, I have found interest in semiconductor fabrication tools as well as machine learning. For these reasons, I want my summer project(s) centered around FPGA, ML, and the SC industry. Outlined below are three ideas I have been playing around with.

trouble-shooting

Trying to Get Signals Agent to Work

11 minute read

Published:

For the past few weeks, I’ve been building a project called Signals Agent, a tool designed to analyze environmental audio using FFTs and other signal processing techniques, powered by an LLM-based reasoning loop. The goal was simple in theory: create an agent that could make sense of ambient sounds by performing multiscale spectral analysis, identify interesting frequency events, and describe likely real-world sources. As with any real-world agentic system, the reality has been much more complex. This blog post documents my recent efforts to improve the accuracy and reliability of the agent, especially in distinguishing between sound sources with overlapping spectral content.