Write and Test Firmware 10X Faster

AssembleAI is the hardware intelligence layer that makes every AI agent an expert on your design.

Hardware Intelligence Demo

Your AI:  I Know Your Hardware Your Firmware Your Schematics Your Chipset Your Requirements Kung Fu

1

Create account and upload files

• Schematics
• Datasheets
2

AssembleAI

• Parses and processes files
• Creates MCP server
3

Your tools know your hardware

• Connect MCP server to your tools
• Done!

See It In Action

Your hardware context, directly in Cursor and Claude Code. No context switching. No manual lookups.

main.c bringup.c sensors.h
Cursor logo Cursor with AssembleAI MCP
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
// Hardware bringup sequence for new board revision
// Board: IoT-Sensor-v2.1 | Date: 2025-01-15
#include "stm32f4xx_hal.h"
#include "board_config.h"
#include "sensors.h"
#include "i2c_util.h"
static uint8_t power_state = 0;
static SPI_HandleTypeDef hspi1;
static I2C_HandleTypeDef hi2c1;
void board_bringup_sequence() {
// Step 1: Configure power rails
HAL_GPIO_WritePin(GPIOC, GPIO_PIN_13, GPIO_PIN_SET); // Enable 3.3V
HAL_Delay(100); // Wait for voltage stabilization
HAL_GPIO_WritePin(GPIOA, GPIO_PIN_8, GPIO_PIN_SET); // Enable 5V rail
HAL_Delay(50);
power_state = 1;
// Step 2: Initialize SPI for temperature sensor
hspi1.Instance = SPI1;
hspi1.Init.Mode = SPI_MODE_MASTER;
hspi1.Init.BaudRatePrescaler = SPI_BAUDRATEPRESCALER_32;
if (HAL_SPI_Init(&hspi1) != HAL_OK) {
Error_Handler();
}
// Step 3: Initialize I2C for IMU and other sensors
hi2c1.Instance = I2C1;
hi2c1.Init.ClockSpeed = 400000; // 400kHz fast mode
if (HAL_I2C_Init(&hi2c1) != HAL_OK) {
Error_Handler();
}
// Step 4: Initialize temperature sensor MAX31855
| // Type to ask AssembleAI about your hardware...
}
⚡
AssembleAI Context Connected
Project Context
Board Version: v2.1
MCU: STM32F407VGT6
Components: 247 loaded
Datasheets: 45 indexed
Intelligent Suggestion
I notice you're bringing up the temperature sensor. Based on your schematic:
MAX31855 Connections (U4):
• PA0 → SO (Pin 3) - MISO
• PA1 → CS (Pin 5) - Chip Select
• PA2 → SCK (Pin 4) - Clock
• R12: 10kΩ pull-up on SO

Initialization sequence:
// MAX31855 Initialization
SPI_HandleTypeDef hspi1;
MAX31855_Init(&hspi1, GPIOA, 1);
HAL_GPIO_WritePin(GPIOA, GPIO_PIN_1, SET);
// CS high initially
HAL_Delay(200); // Startup time
bringup.c • Ln 12, Col 8 • UTF-8 • C
AssembleAI MCP Connected

Direct LLM Integration

Ask ChatGPT, Claude, Gemini, or your favorite LLM about your hardware directly.

ChatGPT ChatGPT
// Ask ChatGPT about YOUR hardware
User: What's the power consumption of my board in sleep mode?
ChatGPT: Based on your schematic, total sleep mode power consumption is approximately 45µA.
• STM32F4 in STOP mode: 10µA
• Voltage regulators quiescent: 25µA
Peripheral consumption breakdown:
• RTC backup: 1µA
• Pull-up resistors: 9µA total
Claude Claude
// Ask Claude about YOUR hardware
User: What's connected to STM32 pin PA0?
Claude: PA0 connects to temperature sensor U4 (MAX31855) through a 10kΩ pull-up resistor (R12).
• Configured as ADC1_IN0 in firmware
• SPI interface on pins PA0-PA2
Key specifications from your design:
• Max temp range: -200°C to +1350°C
• Resolution: 0.25°C
Gemini Gemini
// Ask Gemini about YOUR hardware
User: How do I configure the I2C clock speed for my STM32F4?
Gemini: For STM32F4 I2C, set I2C_CR2.FREQ to your APB1 frequency, then calculate CCR based on your desired speed:
• 100kHz: CCR = APB1_freq / (2 * 100000)
• 400kHz: CCR = APB1_freq / (3 * 400000)
Your specific board uses 84MHz APB1, so:
• 100kHz: I2C_CR2.FREQ = 84, CCR = 420
• 400kHz: I2C_CR2.FREQ = 84, CCR = 70

Built for Hardware Teams

Empowering every engineer to move faster with AI that understands your design.

For Electrical Engineers

For Electrical Engineers

Write your own bringup firmware without waiting for software team availability. Your IDE knows every pin, connection, and constraint.
✓
Instant Bringup Code
Generate initialization sequences for any component on your board, the AI knows your exact connections and voltage rails.
✓
Hardware Validation Tests
Create comprehensive test suites that verify every subsystem, with AI that understands your design constraints and requirements.
✓
Smart Debug Assistant
Troubleshoot signal integrity issues and power sequencing problems with AI that understands your exact schematic and layout.
✓
Catch Issues Early
AI validates your firmware against actual schematics, catching pin mapping errors before hardware is manufactured.
For Firmware Engineers

For Firmware Engineers

Stop drowning in datasheets and schematics. Get instant, accurate answers about your hardware in your favorite editor or IDE.
✓
No More PDF Hell
Never search through 500-page datasheets again - ask questions in plain English and get immediate, accurate answers.
✓
Schematic-Aware Coding
Your AI knows exactly how components are connected - get correct pin mappings, I2C addresses, and SPI configurations instantly.
✓
10x Faster Driver Development
Generate register configurations, timing sequences, and initialization code that's tailored to your exact hardware.
✓
Stay in Your Flow
Access all hardware knowledge directly in VS Code or Cursor - no context switching, no external tools, just answers.

Join teams shipping hardware 10x faster with AI that truly understands their designs.

Works With Your Favorite Tools

One hardware intelligence layer, every AI platform.

AI Models & Assistants

ChatGPT
ChatGPT
OpenAI
Claude
Claude
Anthropic
Google Gemini
Gemini
Google
Grok
Grok
xAI
🦙
Llama
Meta
DeepSeek
DeepSeek
DeepSeek AI

Development Environments

Cursor
Cursor
AI-First IDE
Windsurf
Windsurf
Codeium IDE
GitHub Copilot
Claude Code
Anthropic
VS Code
VS Code
Github Copilot
JetBrains
JetBrains
Cline AI
Powered by Model Context Protocol (MCP)

Compatible with any tool that supports MCP server integration.

Get Started Today

Choose the plan that fits your team's needs.

Monthly
Yearly (save 33%)
Beta Release!

Starter

$0/month

For engineers working on small projects

✓ Run up to 100 MCP calls
✓ Support for 1 hardware project
✓ 100 part datasheets per project
✓ Full integration with Cursor & LLMs
✓ Basic automated schematic parsing

No credit card required, just an email

Pro

$30/month

Professional engineers working on hardware projects

✓ Run unlimited MCP calls
✓ Support for 100 hardware projects
✓ 1000 part datasheets per project
✓ Automated OCR datasheet capture
✓ Full integration with Cursor & LLMs
✓ Advanced netlist extraction & analysis
✓ Native EDA formats (Altium, KiCad)

Join our waitlist for early access

Ultra

$300/month

For advanced engineers building complex products

✓ Unlimited MCP calls
✓ Unlimited hardware projects
✓ Unlimited datasheets with OCR
✓ Native EDA formats (KiCad, Altium)
✓ Expert Reviewed Datasheet Capture
✓ Advanced netlist extraction & analysis
✓ Git integration for firmware repos
✓ Flagship AI model access
✓ Private cloud or on-premise deployment
✓ SSO & advanced security controls
✓ Dedicated support engineer

Email us for full integration into your workflow

Your Hardware, Fully Understood

Complete hardware context at your fingertips, continuously updated from schematics, firmware, and design requirements.

Schematic Intelligence

Automatically extracts netlists, component connections, and design constraints from your EDA files and PDFs.

Datasheet Mining

Indexes thousands of datasheets, making pin mappings, electrical specs, and timing diagrams instantly searchable.

MCP Protocol

Works with any LLM that supports Model Context Protocol - Cursor, Claude Desktop, and more.

Firmware Aware

Links your hardware design to actual firmware implementation, catching mismatches before they happen.

Secure & Local

Your IP stays yours. Run locally or in your private cloud. End-to-end encryption for team sync.

Instant Setup

Drop in your project files and start querying in minutes. Supports KiCad, Altium, Eagle, and more.

Built by Hardware Experts

Our team has worked on products like the Apple iPhone, Google Pixel, Meta Oculus, and more.

Team

Nima Banai

Co-Founder & CEO

Team

Craig Schindler

Co-Founder & CTO

Google Google
Meta Meta
Apple Apple
Amazon Amazon
Team

Join Our Team!

Come build cutting-edge AI tools that will empower hardware engineers.

Apply
Assemble Labs
Assemble Labs Inc. Made in USA 🇺🇸
© Assemble Labs Inc. — All rights reserved.
How It Works Benefits Compatibility Team Contact Us
Get Started

Write and Test Firmware 10X Faster

The hardware intelligence layer that makes every AI agent an expert on your design.

Hardware Intelligence Demo

Your AI:  I Know Your Hardware Your Firmware Your Schematics Your Chipset Your Requirements Kung Fu

1

Create account and upload files

• Schematics
• Datasheets
2

AssembleAI

• Parses and processes files
• Creates MCP server
3

Your tools know your hardware

• Connect MCP server to your tools
• Done!

See It In Action

Your hardware context, directly in Cursor and Claude Code. No context switching. No manual lookups.

main.c bringup.c sensors.h
Cursor logo Cursor with AssembleAI MCP
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
// Hardware bringup sequence for new board revision
// Board: IoT-Sensor-v2.1 | Date: 2025-01-15
#include "stm32f4xx_hal.h"
#include "board_config.h"
void board_bringup_sequence() {
// Step 1: Configure power rails
HAL_GPIO_WritePin(GPIOC, GPIO_PIN_13, GPIO_PIN_SET); // Enable 3.3V
HAL_Delay(100); // Wait for voltage stabilization
// Step 2: Initialize critical sensors
// Temperature sensor MAX31855 on SPI1
| // Type to ask AssembleAI about your hardware...
}
⚡
AssembleAI Context Connected
Project Context
Board Version: v2.1
MCU: STM32F407VGT6
Components: 247 loaded
Datasheets: 45 indexed
Intelligent Suggestion
I notice you're bringing up the temperature sensor. Based on your schematic:
MAX31855 Connections (U4):
• PA0 → SO (Pin 3) - MISO
• PA1 → CS (Pin 5) - Chip Select
• PA2 → SCK (Pin 4) - Clock
• R12: 10kΩ pull-up on SO

Initialization sequence:
// MAX31855 Initialization
SPI_HandleTypeDef hspi1;
MAX31855_Init(&hspi1, GPIOA, 1);
HAL_GPIO_WritePin(GPIOA, GPIO_PIN_1, SET);
// CS high initially
HAL_Delay(200); // Startup time
bringup.c • Ln 12, Col 8 • UTF-8 • C
AssembleAI MCP Connected

Direct LLM Integration

Ask ChatGPT, Claude, Gemini, or your favorite LLM about your hardware directly.

ChatGPT ChatGPT
// Ask ChatGPT about YOUR hardware
User: What's the power consumption of my board in sleep mode?
ChatGPT: Based on your schematic, total sleep mode power consumption is approximately 45µA.
Claude Claude
// Ask Claude about YOUR hardware
User: What's connected to STM32 pin PA0?
Claude: PA0 connects to temperature sensor U4 (MAX31855) through a 10kΩ pull-up resistor (R12).
• Configured as ADC1_IN0 in firmware
• SPI interface on pins PA0-PA2
Gemini Gemini
// Ask Gemini about YOUR hardware
User: How do I configure the I2C clock speed for my STM32F4?
Gemini: For STM32F4 I2C, set I2C_CR2.FREQ to your APB1 frequency, then calculate CCR based on your desired speed.

Built for Hardware Teams

Empowering every engineer to move faster with AI that understands your design.

For Electrical Engineers

For Electrical Engineers

Write your own bringup and test firmware without waiting for software team availability.
✓
Instant Bringup Code
Generate initialization sequences for any component on your board.
✓
Hardware Validation Tests
Create comprehensive test suites that verify every subsystem.
✓
Smart Debug Assistant
Troubleshoot signal integrity issues and power sequencing problems with AI.
✓
Catch Issues Early
Validate your firmware against actual schematics, catching pin mapping errors and more.
For Firmware Engineers

For Firmware Engineers

Stop drowning in datasheets and schematics. Get instant, accurate answers about your hardware in your favorite editor or IDE.
✓
No More PDF Hell
Never search through 500-page datasheets again, get immediate, accurate answers.
✓
Schematic-Aware Coding
Get correct pin mappings, I2C addresses, and SPI configurations instantly.
✓
10x Faster Driver Development
Generate register configurations and initialization code that's tailored to your exact hardware.
✓
Stay in Your Flow
Access all hardware knowledge directly in VS Code or Cursor.

Join teams shipping hardware 10x faster with AI that truly understands their designs.

Works With Your Favorite Tools

One hardware intelligence layer, every AI platform.

AI Models & Assistants

ChatGPT
ChatGPT
OpenAI
Claude
Claude
Anthropic
Google Gemini
Gemini
Google
Grok
Grok
xAI
🦙
Llama
Meta
DeepSeek
DeepSeek
DeepSeek AI

Development Environments

Cursor
Cursor
AI-First IDE
Windsurf
Windsurf
Codeium IDE
GitHub Copilot
Claude Code
Anthropic
VS Code
VS Code
Github Copilot
JetBrains
JetBrains
Cline AI
Powered by Model Context Protocol

Compatible with any tool that supports MCP server integration.

Get Started Today

Choose the plan that fits your team's needs.

Monthly
Yearly (save 33%)
Beta Release!

Starter

$0/month

For engineers working on small projects

✓ Run up to 100 MCP calls
✓ Support for 1 hardware project
✓ 100 part datasheets per project

No credit card required, just an email

Pro

$30/month

Professional engineers working on hardware projects

✓ Run unlimited MCP calls
✓ Support for 100 hardware projects
✓ 1000 part datasheets per project
✓ Native EDA formats (Altium, KiCad)

Join our waitlist for early access

Ultra

$300/month

For advanced engineers building complex products

✓ Unlimited MCP calls
✓ Unlimited hardware projects
✓ Unlimited datasheets with OCR
✓ Native EDA formats (KiCad, Altium)
✓ Expert Reviewed Datasheet Capture
✓ Advanced netlist extraction & analysis
✓ Git integration for firmware repos
✓ Private cloud or on-premise deployment
✓ SSO & advanced security controls

Email us for full integration into your workflow

Your Hardware, Fully Understood

Complete hardware context at your fingertips, continuously updated from schematics, firmware, and design requirements.

Schematic Intelligence

Automatically extracts netlists, component connections, and design constraints from your EDA files and PDFs.

Datasheet Mining

Indexes all your datasheets, making pin mappings, electrical specs, and timing diagrams instantly searchable.

Firmware Aware

Links your hardware design to actual firmware implementation, catching mismatches before they happen.

Built by Hardware Experts

Our team has worked on products like the Apple iPhone, Google Pixel, Meta Oculus, and more.

Team

Nima Banai

Co-Founder & CEO

Team

Craig Schindler

Co-Founder & CTO

Google Google
Meta Meta
Apple Apple
Amazon Amazon

Backed By:

Join Our Team!

Come build cutting-edge AI tools that will empower hardware engineers.

Apply
Assemble Labs
Assemble Labs Inc. Made in USA 🇺🇸
© Assemble Labs Inc. — All rights reserved.