Rapid Nordic IoT Firmware Development

Title:
AI-Assisted Firmware Development for Nordic Thingy91X (nRF5340 + nRF9151) — Full Project in 1 Day

Description:

We are looking for a highly skilled embedded firmware developer who can build the full initial firmware stack for our IoT device based on Nordic Semiconductor's Thingy91X platform. This is a single-day project, expected to be completed within ~10 hours, using AI tools such as Cursor, Claude, or similar for rapid code generation and refactoring.

Hardware Platform:
Board: Nordic Thingy91X

Main SoCs:

nRF5340 (Bluetooth LE Audio / local control)

nRF9151 (LTE-M / cloud comms / file transfer)

PMIC: Nordic NPM1300 (optional drivers ready)

Flash: External 8GB QSPI storage

Firmware Tasks:
Setup and clean full firmware repositories using Nordic Connect SDK / Zephyr examples:

nRF5340 audio example

nRF9151 multi-service example

Board files, DTS, Kconfig

Place all files (approx 300 files, ~30k lines of code) into a single shared Git repository.

Use AI-assisted refactoring (via Cursor, Claude, etc) to:

Clean and reduce unused code

Simplify project structure

Optimize files to ~100 files / ~10k lines

Define and implement cross-core APIs:

SPI or IPC interface between nRF5340 and nRF9151

Downloading LC3 encoded music files (via LTE-M, SoftSIM Onomondo)

Flash storage management

LE Audio streaming to connected Bluetooth headset (JLab JBuds Lux ANC)

Voice command transmission and server roundtrip

Server integration for GPS reporting and FOTA checks

Build & flash firmware to Thingy91X hardware at the end of the session.

Perform basic hardware test of dataflow across both cores.

Target Development Flow:
Shared Git repository — all code visible for AI tooling.

AI-assisted coding — all generation, refactoring, cleanup and documentation should leverage AI where possible.

Total time budget: 8–10 hours.

Budget: $300 USD fixed price.

Ideal Candidate:
Excellent experience with Nordic Semiconductor SDK, Zephyr RTOS, nRF5340 and nRF91x1 platforms.

Familiar with LTE-M, LE Audio, QSPI flash storage.

Fluent in using AI-powered coding tools like Cursor, Claude, GPT-4o or similar.

Efficient and pragmatic — results-oriented.

Note:
You will work independently, but we can provide technical clarifications if needed.

Full project context will be provided upfront with initial source trees, device-tree files and board support packages.

The goal is not "perfect code" but a fully functional starting point firmware that we can extend. The architecture must be clean, functional and easily maintainable.