AI Models
AI models are the core of any intelligent system, determining how effectively tasks are executed and interactions managed. Selecting the right model ensures you achieve an optimal balance of speed, accuracy, and contextual understanding—key for delivering high-quality user experiences and efficient performance. Here are the models we offer and their traits.
The Symphona Platform can also be connected to any custom LLM.
Prompting Basics
What Is AI Prompting?
A prompt is simply the instruction or question you provide to an AI. Think of it as a way to tell the AI exactly what you need—much like asking a friendly colleague for help. By writing a good prompt, you guide the AI to generate a response that is relevant and useful.
Why Is Prompting Important?
Effective prompting is the key to getting accurate and helpful answers from an AI. When you give clear instructions, the AI can focus on the right information and avoid mixing or summarizing details unnecessarily.
How Does It Work?
The AI reads your prompt, uses its training and any connected knowledge base, and then responds based on the guidance you provided. A well-crafted prompt minimizes ambiguity and ensures the AI stays on topic.
The Basics of Writing Prompts
- Keep It Clear and Specific:
- Provide Context:
- Focus on One Question at a Time:
- Use Examples When Needed:
Use simple language and state exactly what you need. Instead of asking, "Tell me about our services," try: “Can you explain the key features of our customer support service?”
Include background information so the AI understands your situation. For example: “I work in sales and need to describe our return policy for new customers.”
Asking multiple questions in one prompt can confuse the AI. Break your questions into individual parts.
If you expect a specific format or style, provide a short example in your prompt. This helps the AI understand your desired output.
Best Practices for Effective Prompts
- Be Direct and Detailed:
- Encourage Specificity:
- Stay on Topic:
- Iterate and Refine:
Clear instructions lead to better answers. Avoid vague language that might cause the AI to guess or summarize too much.
Ask the AI to include exact details from your knowledge base instead of summarizing. For example: “Return the exact text from our policy document.”
Make sure your prompt clearly indicates the subject matter so the AI doesn’t mix unrelated topics.
Don’t worry if your first prompt isn’t perfect. Experiment with phrasing and details until you get the answer you need.
Advanced Prompting Techniques
- Chain-of-Thought Prompts:
- Set a Role or Context:
Guide the AI to think through a problem step-by-step. This can be as simple as: “List the steps needed to troubleshoot a network issue.”
Sometimes, giving the AI a role improves its responses. For example: “You are a customer support assistant for ForgeWorks Industrial Systems. You will…”
Common Pitfalls and How to Avoid Them
- Ambiguous Prompts:
- Overcomplicating Your Prompt:
- Lack of Context:
Vague questions can lead to incomplete or inaccurate answers. Always add enough detail so the AI knows exactly what you need.
Avoid unnecessary complexity. The goal is clarity; keep sentences short and direct.
Without background information, the AI might miss the point of your query. Always include a brief explanation if needed.
Example Prompts
- For a Detailed Explanation:
- For Step-by-Step Guidance:
- For a Quick Summary: “
- For Specific Information: “
“Can you provide a detailed explanation of the benefits of our premium support plan, including response times and service features?”
“List step-by-step instructions on how to reset the password on our customer portal.”
“Summarize the key points of our return policy in one paragraph.”
“Return the exact installation guide for our latest product version as listed in our knowledge base.”