Features

SensAI comes packed with features that let you get more out of every AI conversation. This guide covers each one in plain language, so you know exactly what is available and how to use it.

Model Selection

SensAI gives you access to a curated selection of the world's best AI models across multiple providers. You are never locked into a single model — switch at any time to find the one that suits the task.

Available Providers

Each provider brings its own strengths — OpenAI for broad instruction-following, Anthropic for careful long-context reasoning, Google for factual and multimodal tasks, and Meta for open-weight versatility.

How to Change Models

1. Open any agent (e.g. LLM Farm)
2. Look for the Settings panel — click the settings icon in the agent workspace
3. Find the Model dropdown
4. Select the model you want to use
5. Your next message will use the new model

You can switch models between messages at any time without losing your conversation context.

Choosing the Right Model

• Larger models (GPT-4o, Claude 3.5 Sonnet, Gemini 2.5 Pro) produce more detailed, thoughtful responses. They are better for complex tasks but may be slightly slower.
• Smaller models (GPT-4o Mini, Claude 3.5 Haiku, Gemini 2.0 Flash) respond faster and are ideal for quick questions, summaries, and straightforward tasks.

A good rule of thumb: start with a smaller model for speed, and switch to a larger one if you need deeper analysis or more nuanced answers.

Thinking Mode

Thinking Mode tells the AI to reason through a problem step by step before giving you its final answer. This is particularly useful when accuracy matters more than speed.

What It Does

When Thinking Mode is active, the AI does not jump straight to an answer. Instead, it works through the problem methodically — considering different angles, checking its logic, and only then producing a final response. You can see this reasoning process displayed before the answer, giving you full visibility into how the conclusion was reached.

When to Use Thinking Mode

• Complex math or logic problems
• Code debugging — the AI walks through the issue methodically
• Comparing options — the AI weighs pros and cons before concluding
• Multi-step reasoning where a quick guess might go wrong

How to Enable

1. Open the chat input area in any agent
2. Click the brain icon in the input toolbar
3. Thinking Mode is now active — send your message as normal
4. You will see the AI's reasoning process displayed before its final answer

You can toggle Thinking Mode on or off at any time between messages.

Web Search

Web Search allows the AI to look things up on the internet in real time before answering your question. Without Web Search, the AI relies only on its training data, which has a knowledge cutoff date. With Web Search enabled, it can access current information.

What It Does

When Web Search is active, the AI performs a live search query as part of answering your message. It reads relevant results from the web, synthesizes the information, and incorporates it into its response — giving you answers grounded in up-to-date sources rather than potentially outdated training data.

When to Use Web Search

• Questions about recent events or news
• Current pricing, availability, or product comparisons
• Fact-checking claims that may have changed recently
• Anything where you need up-to-date, accurate data

How to Enable

1. Open the chat input area
2. Click the globe icon in the input toolbar
3. Web Search is now active — send your message
4. The AI will search the web as part of forming its answer

Note: Web Search adds a brief lookup step, so responses may take a moment longer to arrive.

File Attachments

You can upload files directly into the chat and ask the AI questions about them. This is useful when you want the AI to read something specific rather than relying on general knowledge.

Supported File Types

SensAI supports a wide range of common file formats for attachment, including documents, code files, spreadsheets, and plain text files. The AI reads the contents of each file and uses them directly when forming its response.

How to Attach Files

1. Open the chat input area
2. Click the paperclip icon in the input toolbar
3. Select one or more files from your device
4. You can attach up to 5 files per message
5. Type your question or instruction alongside the files and send

The AI will read the contents of your files and use them to answer your question.

Tips for File Attachments

• Be specific in your question — instead of "what is in this file?", try "summarize the key findings from this report" or "find any bugs in this code"
• Combine files — you can attach multiple related files in a single message, for example a spec document and a code file
• Use files as reference — attach a style guide or template and ask the AI to follow it when producing new content

Knowledge Base (Knowledge Agent)

The Knowledge Agent takes file attachments a step further. Instead of uploading files one message at a time, you can build a persistent Knowledge Base — a collection of documents the AI can search any time you ask a question.

Every answer from the Knowledge Agent includes citations, so you can see exactly which document and passage the answer came from.

Supported file types: PDF, DOCX, TXT

Setting Up a Knowledge Base

1. Navigate to Knowledge Agent in the sidebar
2. Click the Knowledge Base tab at the top of the workspace
3. Click Create Knowledge Base and give it a name
4. Click Upload Documents and select your files
5. Wait for processing to complete — this usually takes a few seconds per document
6. Switch to the Chat tab
7. Make sure your Knowledge Base is selected in the settings
8. Ask your question — the AI will search your documents and respond with cited answers

How It Works

When you upload documents, SensAI extracts the text, splits it into searchable chunks, and stores them in a vector database. When you ask a question, the system finds the most relevant chunks from your documents and sends them to the AI alongside your question. This is why the Knowledge Agent can answer precisely and cite sources — it only works from what is actually in your documents.

This approach is especially useful for internal documents, policy manuals, research papers, or any collection of files you want to query regularly.

Understanding Citations

Each answer shows references to the specific chunks of text the AI used. You can click a citation to view the original passage, so you always know where the information came from and can verify it yourself. Citations make the Knowledge Agent trustworthy for high-stakes use cases where you need to trace every claim back to a source.

Chat History

Every conversation you have in SensAI is automatically saved. You never need to copy or export anything to keep a record of your work.

Accessing Your History

Chat history is accessible from the History tab within each agent's workspace. Each agent keeps its own separate history, so your LLM Farm conversations are kept distinct from your Knowledge Agent sessions. Click any past conversation to resume it with full context intact — the AI remembers what was discussed.

Searching Conversations

Use the search bar in the History tab to look for a specific conversation by keyword. Conversations are organized chronologically, making it easy to scroll back to a particular date. If you remember a word or phrase from a past session, search will surface the right conversation quickly.

Tone Settings

Tone settings let you adjust the style and register of AI responses to match your context — whether you need something formal for a client, casual for a quick note, or highly technical for a developer audience.

Available Tones

• Professional — Formal, polished language suited to business communication
• Casual — Relaxed, conversational language for informal contexts
• Technical — Precise, detail-oriented language for technical topics
• Creative — Expressive, imaginative language for writing and brainstorming

Each tone shapes not just word choice but also structure and level of detail, so responses feel genuinely appropriate for the context rather than just slightly rephrased.

How to Change Tone

1. Open the Settings panel within any agent (click the settings icon)
2. Find the Tone option
3. Select your preferred tone
4. All subsequent responses in that session will use the selected tone

You can change the tone at any point in a conversation, not just at the start.

Custom Instructions

Custom Instructions let you give the AI persistent guidance that applies to every conversation — so you do not have to repeat yourself each time.

How to Set Instructions

1. Open the Settings panel within any agent
2. Find the Custom Instructions field
3. Type your instructions in plain language
4. Save — these instructions will now be included in every conversation with that agent

You can update or clear your custom instructions at any time from the same settings panel.

Example Instructions

• "Always respond in bullet points"
• "Focus on healthcare context when relevant"
• "Keep all responses under 200 words"
• "I am a software engineer — assume technical familiarity"
• "Avoid jargon and explain things simply"
• "Always suggest next steps at the end of your response"

Custom Instructions are especially powerful when combined with Tone Settings — for example, setting a Technical tone and adding an instruction like "assume I am familiar with distributed systems" gives the AI a clear picture of exactly how to communicate with you.