Safe Haven AI Transforms Legal Aid Access with 24/7 Voice-Powered Intake Assistant
Discover how Legal Aid Chicago deployed Safe Haven AI’s voice assistant and achieved the equivalent of 3 full-time staff days in just 3 weeks—automatically handling hundreds of family law calls (totaling 22+ hours) while catching critical domestic violence cases that might have otherwise fallen through the cracks
Legal Aid Chicago, serving low-income clients for over 50 years, partnered with Safe Haven AI to tackle an overwhelming demand crisis: over 78,000 calls annually with insufficient staff to do intake or properly refer everyone who needed assistance.
“We receive over 6,500 requests for service by phone every month. Our pilot with Safe Haven showed that AI can keep the lines open, guide callers to the right help, and preserve human capacity for where it’s needed most.”
Overwhelming demand meets limited capacity
Legal Aid Chicago faced a critical access-to-justice bottleneck that affects legal aid organizations nationwide:
- —6,500+ monthly calls flooding intake lines
- —Insufficient staff to handle screening and referrals
- —Thousands of applicants unable to access help due to capacity constraints
- —Staff burnout from constant high-pressure intake work
- —Inconsistent screening leading to missed opportunities and inappropriate referrals
The traditional intake process required extensive human hours for initial screening, qualification assessment, and appointment scheduling—time that could be better spent on direct legal advocacy.
AI-powered intake transformation
Safe Haven AI deployed a sophisticated voice-powered assistant that revolutionized Legal Aid Chicago’s family law intake process. The system seamlessly integrated with their existing phone infrastructure while providing intelligent screening, qualification assessment, and identification of times for potential callbacks.
Intelligent Workflow:
- Initial screening through conversational AI
- Information collection and eligibility assessment using Legal Aid Chicago’s specific criteria
- Automated referrals for non-qualified clients
- Overflow handling - additional eligible callers provide information for potential future contact
- Real-time transcription and case flagging for staff review
- Identification of times for callbacks with intelligent capacity for scheduling integration onto calendars (though not implemented in this short pilot)
Staff Dashboard:
Our AI agent, which automatically integrates into several case management solutions (such as LegalServer), is directly connected to an intuitive platform for the pilot to allow intake staff to:
- —Review eligibility determinations with full reasoning
- —Access complete call transcripts
- —Manage callback scheduling
- —Track which calls had been reviewed
- —Add notes and override AI decisions if they felt it was necessary
Iterative development and refinement
Initial Development: Safe Haven AI worked through multiple iterations of the family law line with Legal Aid Chicago's team, refining the conversational flow and eligibility criteria. Typical implementations require 2-3 iterations, and we were able to communicate about and complete these refinements via email and short calls.
System Integration: The technical integration was straightforward—Safe Haven AI's phone number connected directly to Legal Aid Chicago's system, routing family law calls to the AI assistant. No infrastructure changes were required on Legal Aid Chicago's end.
Pilot: We began our three week pilot on August 11th, 2025. Callers with family law concerns were directed to our voice agent to complete the intake and schedule a follow-up if eligible.
Process Optimization: Halfway through the pilot, Legal Aid Chicago realized their traditional intake was screening too many people as eligible. Safe Haven AI helped redesign both the questions and the entire intake process, creating a more efficient and accurate system.
Dramatic efficiency gains
The pilot program exceeded expectations, demonstrating that thoughtfully designed AI can transform legal aid delivery:
Operational Impact:
- —417 total calls triaged during pilot period (family law only, over 3 weeks)
- —281 screened out
- —136 screened in for immediate review for additional services
- —22+ hours of client engagement handled automatically
- —3.24 minutes average call length for efficient screening
The AI talked to hundreds of people who otherwise would have been told the queue was closed. As a result, hundreds more callers were able to get useful information or referrals from the call, even if they were ineligible for services.
Efficiency Gains:
- —44 priority intakes for quick connection to intake staff for domestic violence survivors seeking help with divorce or custody-related legal issues
- —200+ referrals for non-priority clients to appropriate organizations
- —20 specialized handling for urgent cases (court dates within a week)
Unexpected Lessons:
- —Redesign of screening questions achieved greater efficiency in the screening process
Equivalent of hiring additional full-time staff
The pilot handled 22 hours of phone calls—equivalent to having an extra full-time intake worker for three entire days, even just focusing solely on family law. This represents significant operational value:
Resource Multiplication:
- —3+ days of human labor automated
- —Consistent 24/7 availability without overtime costs
- —Exponentially increased capacity given that the AI can handle volume spikes and up to 1,000 calls at any given moment
- —Standardization across all intake interactions
Staff Satisfaction: Rather than replacing human workers, the system frees legal advocates to focus on complex intakes and cases requiring human expertise and empathy.
From Legal Aid Chicago: Safe Haven AI did a remarkably good job of screening for priority calls.
With traditional phone triage, 46% of family law-related calls were ultimately rejected as out-of-priority after intake staff had gathered more information from the caller. By contrast, the Safe Haven AI agent reduced that rate to just 21%—a striking improvement in identifying priority legal issues upfront.
Client Comfort with AI Disclosure
- —People were surprisingly comfortable describing sensitive situations to the AI agent
- —This contradicts common assumptions that people won’t share personal/traumatic details with technology
- —Clients seemed more willing to disclose abuse details to AI than expected
According to Legal Aid Chicago—s follow-up survey, caller experience was overwhelmingly positive. The average rating was 4.1(5=best). Almost 55% of users surveyed rated their experience as 5 out of 5, and 24% rated it 4 out of 5. Here are some of the comments shared by callers:
It was clear and everything was understood, very easy.
Easy to speak to and prompt response.
Short and simple conversation, pleasant.
Found it easy to speak to the AI assistant, just like a live person.
Clear, and my voice was understood and I did not have to repeat myself.
Best service.
Transforming client experiences
Domestic Violence Priority: DV survivors received expedited screening and priority scheduling, eliminating dangerous wait times for critical legal assistance. The AI agent helped uncover domestic violence cases that weren—t initially flagged through the phone menu. Structured AI questioning revealed high-risk situations that callers didn—t self-identify via menu options. In one case, the voice agent identified a high-risk domestic violence situation involving stalking and concerns about retaliation that might have otherwise fallen through the cracks during busy periods. The caller was immediately flagged for urgent attention and received a same-day callback, potentially preventing escalation of violence.
Continue the impact you just read about
Explore a tailored pilot, walk through our full intake stack, or connect with our team about scaling AI intake across your organization.