Introduction

Overview

The arrival of practical, language-based AI assistants marks a major turning point in how professionals interact with complex software. In just a few years, large language models (LLMs) have evolved from academic curiosities into powerful tools capable of understanding context, answering domain-specific questions, and helping users complete real tasks. For engineers this represents an opportunity to integrate intelligent assistance directly into the design workflow, reducing friction and accelerating results.

As the capabilities of AI models continue to improve, their cost and accessibility have improved just as rapidly. Cloud-based providers offer free tiers and pay-as-you-go options, while open-source models can now run on local hardware. This makes it possible for almost any engineering team, or even an individual user, to benefit from advanced AI without significant infrastructure investment.

More than just a convenience, AI assistants are becoming a reliable productivity multiplier. Whether you are troubleshooting a simulation, looking up a component function, or writing embedded code, the ability to ask questions and get meaningful, context-aware answers directly within your design tool is transformative. In many ways it is like having your own junior engineer on standby: always ready, never tired, and surprisingly capable.

In response to this shift, the Proteus Design Suite includes a built-in AI assistant designed specifically for engineering work. This tool integrates tightly with your schematic, simulation data, and embedded code environment to help you design, debug, and understand electronic systems faster and more effectively.

Introducing EDAi

Engineering is a particularly strong candidate for AI assistance. Unlike creative writing or abstract reasoning tasks, circuit design and firmware development tend to follow defined patterns, logical structures, and standardised techniques. This makes the domain well suited to AI reasoning, especially when the assistant has access to real-time context such as your schematic layout or component connections.

To support both professional engineers and educational users, Proteus provides two distinct AI tools: ProPilot, aimed at commercial users, and ProTutor, tailored for learning environments. While both are built on the same foundation, ProTutor is intentionally designed to guide students toward understanding rather than simply supplying answers, encouraging exploration and learning.

Behind the scenes we have partnered with a range of leading AI model providers, including Anthropic (Claude), Open AI (ChatGPT), DeepSeek, Google (Gemini) and Moonshot (Kimi), alongside local model platforms such as Ollama and support for custom endpoints. You can choose the model that best suits your preferences for accuracy, speed, privacy, or cost.

Most importantly, these assistants are embedded directly into the Proteus application. That means you do not need to switch tools or leave your design environment to ask a question, explore an idea, or resolve a problem. With context-aware support always available, EDAi brings real engineering productivity to your fingertips.

The integrated EDAi chat window is available on the schematic capture and VSM Studio modules. Due to technical limitations with the AI models it is not yet enabled directly within the Proteus PCB Layout tool. This is a very fast moving field, we are confident in AI model progression over the next year or so, and we intend to extend capabilities in Proteus when it makes sense to do so.

Top Tips

Learning how to drive an AI model is a skill in itself. The following tips will help you guide the models to providing the best results.

Be specific to get better answers

  • State your goal clearly and give relevant context.
  • Include examples, code, or text you want improved.
  • State LM358 op-amp instead of just op-amp.
  • Provide pin numbers for IC connections.

Always ask smart, reasoning based questions

Use follow on prompts like:

  • "Are you sure?"
  • "Explain the logic behind that."

This helps the AI clarify and improve its answers.

Understand your provider's rate limits

  • AI providers apply rate limits to API accounts, and those limits affect performance.
  • A higher limit makes for smoother interaction, particularly on large designs.
  • Your provider's account or billing pages show the limits that apply to you and what is required to raise them.

If it stalls, rephrase and retry

  • Change your wording or ask a smaller question.
  • Be explicit, because vague prompts lead to vague replies.
  • Ask it for more information or justification.
  • Try "Please re-check your answer".

Be aware of tokens and chat length

  • Long chats use more tokens and may lose context.
  • Watch the token counter at the top right of the chat.
  • Restart if replies get vague or off-track.
  • Keep chats focused on a single task.

You are still the engineer

  • AI can boost productivity, but you are the expert.
  • LLMs are powerful and improving quickly; they are not perfect, so always double-check.
  • Always review results, ask follow-ups, and guide the direction.
  • You are the engineer. ProPilot helps, but you decide.

Examples

Vague prompt "Are my IC's all powered?" Specific prompt "Check that all my ICs are connected to power nets"
Vague prompt "My 555 timer isn't working properly" Specific prompt "My 555 timer is in astable mode. Can you help me calculate the output frequency given R1=10KOhms, R2=47KOhms and C=100nF"

It is also worth understanding a little about how AI chat works, because that helps you understand how best to use it. We strongly recommend you review the best practice guide before working with ProPilot. It is important to understand how to drive the AI models optimally and how to check, confirm and iterate conversations for the best results.

See Also:

What ProPilot Can Do

Model Setup