FAQ and Troubleshooting

Answers to common questions about SensAI and tips for getting the most out of the platform.

SensAI is an AI Agent Studio built by JMR Infotech. It brings together a collection of specialized AI assistants — each designed for a different kind of task — in a single, easy-to-use platform. You can chat with powerful AI models, query your own documents, generate specification documents, summarize reports, draft proposals, and visualize code issues. All from your browser, with no technical setup required.

Do I need any technical skills to use SensAI?

No. SensAI is designed for everyone, regardless of technical background. You do not need to write code, manage API keys, or understand how AI models work. If you can type a question, you can use SensAI. The interface is guided, and every agent includes built-in tours to walk you through the features the first time you use them.

Which AI models are available?

SensAI gives you access to models from multiple leading providers:

OpenAI — GPT-4o, GPT-4o Mini
Anthropic — Claude 3.5 Sonnet, Claude 3.5 Haiku
Google — Gemini 2.0 Flash, Gemini 2.5 Pro
Meta — Llama 4 Maverick, Llama 4 Scout
And others, updated regularly as new models become available

You can switch between models at any time from the Settings panel inside any agent.

Is my data private?

Yes. Your data is private to your account. Knowledge Bases, chat history, uploaded files, and settings are all associated with your individual account and are never shared with other users. SensAI uses session-based authentication, and each account is fully isolated. See the How SensAI Works page for more detail on the security model.

Using SensAI

How do I switch AI models?

Open any agent (e.g. LLM Farm)
Click the settings icon in the agent workspace to open the Settings panel
Find the Model dropdown
Select the model you want
Your next message will use the new model

What file types can I upload?

For the Knowledge Base (Knowledge Agent): PDF, DOCX, and TXT files are supported. You can upload multiple documents to build a searchable Knowledge Base.
For chat attachments (any agent): You can attach up to 5 files per message using the paperclip icon in the chat input. Use this for one-off analysis — summarizing a document, reviewing code, or referencing a file in a conversation.

How does the Knowledge Agent find information in my documents?

The Knowledge Agent uses a technique called Retrieval-Augmented Generation (RAG). Here is the short version:

When you upload documents, they are split into small sections and converted into numerical representations (embeddings) that capture their meaning
These are stored in a searchable index
When you ask a question, your question is also converted into an embedding
The system finds the document sections whose meaning is most similar to your question
Those sections, along with your question, are sent to the AI model, which writes an answer and cites the sources

This is why you see citations in every answer — the agent always knows which part of which document it used.

What is Thinking Mode?

Thinking Mode tells the AI to reason through a problem step by step before giving you its final answer. When enabled, you will see the AI's reasoning process displayed before the conclusion.

It is especially useful for:

Complex math or logic problems
Code debugging
Comparing options or making decisions
Any question where accuracy matters more than speed

Enable it by clicking the brain icon in the chat input toolbar.

What does Web Search do?

Web Search allows the AI to search the internet for current information before answering your question. Without it, the AI uses only its training data, which has a knowledge cutoff date. With Web Search enabled, it can access real-time information.

Use it for recent events, current pricing, fact-checking, or anything where up-to-date information matters. Enable it by clicking the globe icon in the chat input toolbar.

Troubleshooting

The response seems cut off or incomplete

This can happen when a question is very broad or the answer requires more context than the model processed. Try these steps:

Ask a more specific question — narrow down exactly what you need rather than asking for everything at once
Switch to a larger model — models like GPT-4o or Claude 3.5 Sonnet handle longer, more complex responses better
Break it into parts — if you need a long answer, ask in multiple focused questions rather than one large one

The Knowledge Agent is not finding information I uploaded

A few things to check:

Confirm the upload completed — go to the Knowledge Base tab and verify your documents appear there with a completed status
Rephrase your question — try using different words or a more direct question. The retrieval system matches meaning, so paraphrasing can help.
Check the document content — if the information you are looking for is in a scanned image (not selectable text), the text extraction may not have captured it
Lower the similarity threshold — if your settings allow it, reducing the similarity threshold will return a broader set of passages

The tour guide tooltips are not showing up

The guided tour auto-detects elements on the page before showing tooltips. If a tooltip does not appear:

Refresh the page — this resets the tour detection and is usually all that is needed
Wait a moment — if the page is still loading, the tour may start once all elements are ready

If the issue persists, you can restart the tour from the help menu within the agent.

Responses are slow

Response speed depends on the model you are using and current demand. If speed is a priority:

Switch to a smaller model — GPT-4o Mini, Claude 3.5 Haiku, and Gemini 2.0 Flash are optimized for fast responses
Disable Thinking Mode — extended reasoning adds processing time
Disable Web Search — real-time search adds a lookup step before the AI responds