Opens in a new tab

The Cause Outside Your Data: Web Search in Ask Spotlight

Ask Spotlight can now look up public information on the web when you ask it to, so it can explain the things your own data can't.

What’s web search in Ask Spotlight?

Your contacts can tell you that 300 people got in touch about a diverted flight this morning. The reason for the diversion lives outside EdgeTier, in a news report about fog closing the airport. Same story when players complain about how a wager was graded: your data has the anger, and the match result that explains it is somewhere on the web.

Ask Spotlight can now go and look. Ask it to check the web and it runs a search, reads the results, and folds what it finds into the answer alongside your interaction data.

What’s new

It only searches when you ask. You have to request it in the message: “search the web”, “can you look that up”, “check what happened with X”. Without that, Ask Spotlight answers from your data alone, even for a question a search would obviously improve. That keeps the behaviour predictable. When an answer contains anything from outside EdgeTier, you asked for it, and you can see where it came from.

Every result is cited. Anything Ask Spotlight takes from a search comes back with the source and a link. A snippet from a news site is worth a different amount to a number out of your own database, and seeing where a claim came from lets you judge it.

Already using EdgeTier over MCP? You can do the same thing there. If you’ve connected Claude, ChatGPT or another assistant to the EdgeTier MCP server, that assistant brings its own web search, so it can pull your interaction data and public context into one answer without waiting on this flag. The difference is whose search you’re using and what it costs: Ask Spotlight’s runs inside AWS under the controls described below, while an MCP-connected assistant searches through its own provider under your agreement with them.

Example prompts

  • “We had a spike in contacts about flight FR2143 this morning, can you search the web for what happened?” You get the spike quantified from your data and the public cause in the same answer: 312 contacts, four times baseline, and Dublin Airport suspending departures for 90 minutes in fog.
  • “Are there any Premier League matches on this Friday? We get an influx of volume around football.” Useful for staffing a shift when your volume tracks an external calendar you don’t hold anywhere in EdgeTier.
  • “Can you search the web and tell me about the big fashion events last month that could have impacted my volume?” Retail teams can line a contact spike up against a sale event, a drop, or a competitor’s launch.
  • “A lot of players are contacting us about a wager on the Arsenal game being graded wrong, can you check the result?” The match went to extra time, which is usually the whole explanation, and it’s the one fact your interactions don’t contain.

A few things worth knowing

⭐ Searches are capped at three ‘web search requests’ per message, with a maximum of 7 results per query. That’s enough to establish what happened while keeping the answer readable.

⭐ Ask Spotlight is still an analytics assistant. It won’t open a link you paste in, browse a page you send it, or work as a general research tool. The search is there to explain your customer contacts.

If you have any queries about using this within Ask Spotlight, or you’d like it turned on for your organisation, reach out to Customer Success.

Customer-Focused Leaders Trust EdgeTier

  • Electric Ireland Logo

    "We thought at the time that we were putting the customer at the fore. We thought we were doing things right. But in hindsight, we really weren’t because we had no real-time insights whatsoever into customer issues."

  • novibet

    "It has reduced the time for the quality assurance process as it provides clear data and a very robust direction on where to look and what matters the most."

  • EdgeTier Assets - Abercrombie Logo

    "The anomaly feature is a game changer for us. It’s highly accurate and has helped us identify customer issues, agent errors, and even fraud that would have taken us longer to catch."

Employees avatar purple
Employees avatar yellow
Employees avatar blue

Ready to see results?

Let us help your company go from reactive to proactive customer support.

Unlock AI Insights