About
I work at the boundary between hardware and cloud, where a sensor's signal becomes a data pipeline and a control loop becomes an infrastructure decision.
My background spans real-time C/C++ on microcontrollers and FPGAs, through to Python backends, CI/CD pipelines, and AWS cloud infrastructure, built during an MSc in Embedded Systems & IoT and a role as an Information Systems Engineer at Lumen Technologies.
Currently building a portfolio of IoT + AWS projects focused on smart grid and energy-transition applications, combining embedded sensing with serverless, event-driven cloud architecture.
Projects
Smart Grid Energy Monitor
IN PROGRESSTarget architecture: a simulated smart-meter fleet streaming to AWS IoT Core → Kinesis → Timestream, with a live dashboard and SNS anomaly alerting, deployed via Terraform. My target-sector flagship project — not yet at a public-repo stage.
What's built so far: · What's next:
PID Core (C99 Control Library)
SIMULATED + CI TESTEDDependency-free C99 PID controller with anti-windup and a filtered derivative-on-measurement term, unit tested in CI against a DC motor simulator and cross-compiled unchanged for a Cortex-M target.
→ github.com/arfaali-naikar/pid-coreEdge AI Anomaly Detection
SIMULATED + CI TESTEDTinyML vibration/speed anomaly detector for motor control systems: autoencoder trained and evaluated end-to-end in CI, with an edge inference loop bench-tested on an Arduino Nano 33 BLE and retraining synced through AWS SageMaker.
Results are measured on simulated vibration data; the firmware has not been run against a real motor.
→ github.com/arfaali-naikar/edge-motor-anomalyFPGA Audio Signal Processing (FIR Filter)
BENCH TESTED8-tap FIR audio filter for a DE1-SoC Cyclone V FPGA: codec init, serial-to-parallel conversion and filtering in VHDL, recovered from a 2024 coursework report and verified in GHDL simulation. The reconstructed top-level has not been re-verified on physical hardware.
→ github.com/arfaali-naikar/fpga-fir-audio-filterClosed-Loop Motor Control (M2M + PID)
HARDWARE VALIDATEDReal-time DC motor speed regulation on Raspberry Pi using a PID algorithm with encoder feedback, holding a 2Hz target speed across multihomed LAN routing.
→ github.com/arfaali-naikar/closed-loop-motor-control-m2m-pidShow all projects
RSA-Based Multimedia Encryption
BENCH TESTEDRSA cryptography engine in MATLAB for secure text, image, audio and video transmission, using modular exponentiation and signal-distortion analysis under noisy channels.
→ github.com/arfaali-naikar/rsa-multimedia-encryptionSkills
Embedded & Firmware
- C/C++
- VHDL
- Real-Time Systems
- UART
- STM32
Cloud & AWS
- S3
- Terraform
- Amplify
- IoT Core
- Lambda
- DynamoDB
DevOps
- Docker
- Jenkins
- GitHub Actions
- Grafana
- Kubernetes
Data & Backend
- Python
- Django
- Flask
- SQL
- Power BI
Experience
Contributing to software development work alongside full-time employment, building on backend and systems experience from Lumen Technologies.
Previously Operations Team Member, then Customer Experience Advisor. Customer-facing support in a high-volume operations environment.
- Built backend systems and data pipelines with Python, Django and Flask.
- Implemented CI/CD (Jenkins, Docker) and system automation with BMC Control-M.
- Managed monitoring (Nagios, Grafana) and performed vulnerability/penetration testing.
Designed and prototyped secure embedded systems for IoT projects, integrating hardware and software components.
AI Use
AI is a tool I use to build and ship faster, not a substitute for engineering. Across these projects, some portion of the code, including boilerplate, syntax, and scaffolding, is AI-generated or AI-assisted. But the architecture, the logic, the problem-solving, and the ideas behind each project are solely mine, and I hold the result to the same standard as if I'd written every line myself.
Learn more
I treat AI as a guiding tool and assistant, in the way I'd use a compiler, a linter, or a senior colleague to review a design. It speeds up execution, but it doesn't make the decisions. The design intent, the trade-offs, and the "why" behind how a system is built come from me first.
This site is a direct example: I'm an embedded systems engineer, not a visual designer, so the layout, styling and theme of this page were AI-generated. The projects it documents were not.