AI, Automation Improves ICE Call Center Efficiency

Immigration and Customs Enforcement (ICE) relies on a special call center — the Homeland Security Investigations Tip Line — for leads on human trafficking, terrorism, cybersecurity crimes and money laundering, but it faces operational challenges due to outdated IT infrastructure.

To remedy this, the call center is exploring artificial intelligence and automation to modernize operations and also relieve human workers from rote tasks, which is part of a growing federal trend to modernize call and contact center IT infrastructure, HSI Tip Line Unit Chief Jody Fasenmyer told GovernmentCIO Media & Research.

Anyone can call the HSI Tip Line to offer tips or information, but tipsters can also fill out the HSI Tip Line’s online form at ice.gov/tips.

“We’re exploring AI processes that read web tips through our online tip form,” Fasenmyer told GovernmentCIO Media & Research. “That technology should also help us categorize those tips based on the tip type. That helps our analysts reduce the amount of time required to process those tips. That also allows us to identify the higher priority leads.”

Tip line callers are queued by priority, which is usually determined by time sensitivity. Callers who may be sex-trafficking victims, for example, are bumped up in the queue.

“Callers select an alleged crime they’re reporting from an automated menu,” Fasenmyer said. “Those calls are prioritized based on what they’re reporting. They’re queued for our personnel based on the experience needed and difficulty level. Time-sensitive calls are given top priority in our system. Those quickly go to the top of our queue so they’re answered very quickly. We provide interpreter support as needed. We can initiate a three-way call and access interpreters for almost 250 languages and dialects.”

Before the pandemic, the HSI Tip Line processed upward of 3,000 phone calls and 2,500 online tips per week. Those numbers dropped when lockdown orders went into effect, but Fasenmyer expects them to tick back up again as commerce and travel resume through 2021.

A big challenge facing the tip line is the lack of IT infrastructure.

“What’s unique about our operation is we don’t currently use any systems or applications that were designed for tip-management purposes,” Fasenmyer said. “They weren’t based on requirements for the unique tip operation. We’re using tools that are available to our analysts that are very helpful for a variety of things, but they weren’t necessarily designed for what we do.”

The tip line has been working with the HSI Innovation Lab to develop new tools and processes for improving the tip line’s efficiency. They’re exploring ways to standardize web forms, automate dissemination of tip reports to ICE field agents and centralize data flows.

“The challenges we run into is, because we’re using tools that weren’t designed for our operation, we’re engaged in very manual processes sometimes,” Fasenmyer said. “Getting metrics on our performance or the outcomes of our tips, that’s a very manual process and requires a lot of time. With the HSI Innovation Lab, we’re looking to automate that process. In processing tip information, there are a lot of extra steps our analysts go through to move information from Word documents into systems while compiling a report. We’re really looking to eliminate a lot of that redundancy and extra steps. Our big focus is on shaving time off (of that) so people aren’t waiting (and get back on the phone with callers).”

Fasenmyer wants a centralized dashboard where analysts and supervisors can “work more easily together” when compiling important tip information for ICE field agents to act upon.

“That replaces a very manual process with shared Google Drive folders and accessing documents,” she said. “That’s really going to help our operations.”

To improve productivity, the HSI Tip Line is building an automation tool to easily search existing data and identify links between different tips, like addresses, names or even topics.

“As we’re inputting information into the system, it’s hopefully pulling up other related things,” she said. “We’re also working to automate how we track and visualize [those] metrics.”

The tip line is also working on AI tools to read the online tip form submissions and translate online tips submitted in different languages.

“Our biggest concern is customer service,” Fasenmyer said. “People want to provide us with leads about alleged illicit activity, and those leads are very important for us to process. Any modernization we’re undertaking, we want to reduce the amount of time we review and process and triage those tips and be able to focus more on getting back to that next call or web tip.”

The HSI Tip Line took an opportunity to explore AI, automation and new software tools to improve operations over the past year due to the pandemic-induced drop in call and online tip volume.

“What’s really working well for us is the pandemic gave us the opportunity to explore some new technologies that allow us to handle phone calls, web tip traffic, from remote locations,” Fasenmyer said. “That was our biggest development recently. We can handle operations 100% remotely. That’s probably our biggest success story.”