Ask Lightspeed Gets Even Smarter

Ask Lightspeed has always been your team’s command center for charitable gaming operations — drafting content, searching your Knowledge Base, looking up Shopify orders, checking the Runway calendar, and analyzing data. Today it gets a major upgrade: the ability to search the entire web, visualize data with interactive charts, and deliver it all faster.

Web Search: Verified Answers with Sources

This is the big one. Ask Lightspeed can now search the web in real time and bring back current, verified information — complete with clickable source links so you can see exactly where the information came from.

Toggle the Web button next to the model selector and Ask Lightspeed gains access to the open internet. Ask it anything that goes beyond your internal data:

• “What are the latest AGCO regulations for charitable gaming in Ontario?”
• “Who has the largest Catch the Ace raffle in Canada?”
• “What are best practices for promoting a 50/50 raffle on social media?”
• “What did the Ontario Lottery and Gaming Corporation announce this month?”

Instead of guessing or giving you a generic answer from its training data, Lightspeed searches the web first and builds its response from what it actually finds. Every answer includes source links displayed below the response, so you can verify the information yourself or share the original sources with your team.

How It Works

When Web is toggled on, Ask Lightspeed treats every factual question as a web search first. It queries multiple sources, synthesizes the results, and presents a clear answer with citations. Internal tools always take priority — if the answer is in your Knowledge Base, calendar, Shopify data, or Home Base, Lightspeed pulls from there instead. Web search fills the gap for everything else: industry news, regulatory updates, competitor research, best practices, and general knowledge questions.

When Web is toggled off, nothing changes. Ask Lightspeed works exactly as it always has, drawing from your connected tools and internal data. Web search is entirely opt-in, entirely under your control.

Smarter Follow-Up Suggestions

When you use web search, Ask Lightspeed now suggests contextual follow-up searches based on what you just asked. These aren’t generic suggestions — they’re generated by AI based on the specific topic you’re researching, so they actually make sense as a next step. Click one and it fires off a new web search instantly.

Inline Charts and Graphs

Ask Lightspeed can now render interactive charts directly in the conversation. When you ask about data — sales velocity from Heartbeat, Shopify order trends, analysis results, or even statistics from a web search — Lightspeed can visualize it as a bar chart, line graph, pie chart, doughnut chart, or horizontal bar chart, right inline.

You don’t need to ask for a chart explicitly (though you can). When the data lends itself to a visual, Lightspeed will render one proactively. Try prompts like:

• “Show me our sales velocity across all time windows”
• “Chart the ticket tier breakdown”
• “Visualize our revenue trend over the past week”
• “Graph the top 10 cities by ticket sales”

Charts appear inline in the conversation with a title, color-coded datasets, and interactive tooltips on hover. Lightspeed adds text commentary alongside the chart so you get both the visual and the analysis in one response.

Faster Responses with Prompt Caching

Under the hood, we’ve implemented prompt caching across Ask Lightspeed’s API calls. The technical details: the static system prompt, tool definitions, and conversation history are now cached between requests, which means the AI doesn’t have to re-process thousands of tokens of context on every message. In multi-turn conversations, this translates to noticeably faster response times — especially as conversations grow longer.

You won’t see a toggle or a setting for this. It just works. Every Ask Lightspeed conversation is faster now.

Model Selector Update

We’ve streamlined the model selector to two options: Sonnet 4.6 (the default, fast and capable) and Opus 4.6 (the most powerful, for complex tasks). The Haiku option has been removed from the chat interface to keep things simple — Lightspeed still uses Haiku behind the scenes for internal tasks like generating follow-up suggestions.