Skip to content
View rameyjm7's full-sized avatar
2 followers · 1 following
  • Richmond, VA
  • 19:32 (UTC -04:00)
  • LinkedIn in/rameyjm

Achievements

Achievement: Pull SharkAchievement: Arctic Code Vault Contributor

Achievements

Achievement: Pull SharkAchievement: Arctic Code Vault Contributor

Block or report rameyjm7

Block user

Prevent this user from interacting with your repositories and sending you notifications. Learn more about blocking users.

You must be logged in to block users.

Maximum 250 characters. Please don’t include any personal information such as legal names or email addresses. Markdown is supported. This note will only be visible to you.
Report abuse

Contact GitHub support about this user’s behavior. Learn more about reporting abuse.

Report abuse
rameyjm7/README.md

Jacob M. Ramey

Senior systems engineer focused on embedded AI, RF/sensor intelligence, SDR platforms, GPU-ready inference, and connected-device software.

I build the software layer around real systems: capture pipelines, edge compute, dashboards, model evaluation, deployment tooling, telemetry, cybersecurity-aware platform work, and Linux/Windows IoT device software.

Portfolio

  • LinkedIn - background, roles, and professional context
  • Hugging Face - public model and ML artifacts

Selected Proof Points

  • Built RF and sensor-intelligence systems spanning SDR capture, passive discovery, operator dashboards, multi-band wireless survey evidence, and field-oriented prototype workflows.
  • Built connected-device software for Linux and Windows IoT systems, including telemetry, diagnostics, secure-update support, platform hardening, and production-image workflows.
  • Supported medical-device engineering evidence for 510(k)-cleared product environments, cybersecurity review, embedded platforms, and regulated technical documentation.
  • Built a production-style AI inference service with FastAPI, ONNX-ready runtime loading, deterministic fallback inference, Prometheus metrics, Docker, benchmarks, tests, and CI.
  • Built LLM adaptation and evaluation workflows covering LoRA, QLoRA, SFT, activation-level unlearning, prompt perturbation analysis, and reasoning benchmarks.
  • Built RF/IQ ML pipelines with live SDR replay/receive, ONNX export, TensorRT deployment on NVIDIA Jetson, and Nsight-profiled edge inference.
  • Built data and ML workflows using Kafka, S3, Airflow, TensorFlow, Docker, and AWS.

Current Focus

  • RF and sensor intelligence: SDR capture, IQ pipelines, spectrum monitoring, passive discovery, and operator-facing RF dashboards
  • Embedded AI: NVIDIA Jetson, ONNX, TensorRT, CUDA-enabled inference, edge deployment, and profiling
  • Connected devices: Linux services, telemetry, diagnostics, secure updates, field reliability, and production images
  • Medical-device platforms: 510(k) cybersecurity evidence support, Linux and Windows IoT devices, secure platform hardening, and regulated engineering documentation
  • LLM and AI evaluation: LoRA/SFT workflows, reasoning benchmarks, inference services, and reproducible ML pipelines

Public Systems And Evidence

RF / SDR Systems

rf-signal-intelligence Real-time RF drone-classification pipeline covering public RF datasets, live SDR replay/receive, ONNX export, Jetson TensorRT FP16 deployment, and Nsight/trtexec profiling. Includes public result cards and links to system-level evidence.

sdr-shark Applied RF/ML platform with React UI, Python backend, SDR streaming, scanner workflows, decoder plugins, signal analysis dialogs, waterfall visualization, and integration points for RF/IQ classifiers.

rf-sentinel Multi-SDR RF intelligence platform for passive discovery and normalized event tracking across Bluetooth, BLE, Zigbee, Wi-Fi, TPMS, FM, cellular, and LF/MF signals.

Fielded Proof Projects

AirScope Multi-band wireless assurance platform for 900 MHz HaLow and 2.4 / 5 / 6 GHz Wi-Fi survey evidence, inventory, channel analysis, and report workflows.

PASSIVE-SHIELD Prototype evidence chain for passive RF sensing, edge-node status, acoustic corroboration, event sharing, track lifecycle, and operator cues.

AI / Inference Engineering

gpu-inference-pipeline Production-style AI inference service with FastAPI, ONNX-ready runtime loading, deterministic fallback inference, Prometheus metrics, Docker deployment, benchmark tooling, tests, and CI.

qwen-sft-reasoning-benchmark Supervised fine-tuning and evaluation workflow for Qwen2.5-3B-Instruct using LLaMA-Factory, Hugging Face tooling, benchmark evaluation, and reasoning-task analysis.

masked-emotion-lora-benchmark Reproducible LoRA adaptation benchmark for masked-emotion reasoning across encoder and generative model families.

Technical Stack

Python, C, C++, JavaScript, React, Flask, FastAPI, Linux, Windows IoT, Docker, systemd, MQTT, WebSocket streaming, SoapySDR, HackRF, bladeRF, RTL-SDR, NVIDIA Jetson, CUDA, ONNX, TensorRT, TensorFlow, PyTorch, Hugging Face, OpenCV, embedded Linux, QNX, PetaLinux, NXP i.MX, Xilinx Zynq UltraScale+ MPSoCs, FPGA platform integration.

Contact

LinkedIn: https://www.linkedin.com/in/rameyjm
Hugging Face: https://huggingface.co/rameyjm7

Pinned Loading

  1. gpu-inference-pipeline gpu-inference-pipeline Public

    Production-style AI inference service with ONNX-ready runtime, FastAPI, metrics, Docker, benchmarks, tests, and CI.

    Python

  2. llm-preference-unlearning llm-preference-unlearning Public

    Research prototype exploring activation-level unlearning methods to improve robustness and alignment in large language model–based recommender systems.

    Jupyter Notebook

  3. qwen-sft-reasoning-benchmark qwen-sft-reasoning-benchmark Public

    Supervised fine-tuning and benchmark evaluation workflow for Qwen2.5-3B-Instruct reasoning performance.

    Shell

  4. masked-emotion-lora-benchmark masked-emotion-lora-benchmark Public

    Reproducible masked-emotion benchmark with baseline reproduction and LoRA adaptation across encoder and generative model families.

    Jupyter Notebook

  5. rf-signal-intelligence rf-signal-intelligence Public

    Real-time RF signal intelligence pipeline for drone RF classification, live SDR replay/receive, ONNX export, TensorRT deployment on NVIDIA Jetson, and Nsight-profiled edge inference.

    Python 6 2

  6. sdr-shark sdr-shark Public

    Personal portfolio copy of SDR-Shark, an RF signal-intelligence and SDR visualization platform demonstrating live spectrum monitoring, decoder integration, RF activity tracking, and ML integration.

    Python 1