Key Metric
3x service capacity
The Context
A regional office of the Defensoría del Pueblo (Office of the Ombudsman) in Lima, Perú, serving citizens with inquiries about fundamental rights, public services, and access to justice. Annual budget of approximately $800,000 USD, funded by the Peruvian state.
Practice Area: Citizens' rights defense — access to healthcare, education, public services, gender-based violence, and labor rights
Jurisdiction: Perú (Metropolitan Lima and Callao)
Team Size: 6 staff attorneys, 3 legal assistants, 8 university interns, 4 receptionists
The Challenge
Problem: The office was turning away approximately 65% of citizen inquiries due to capacity limitations. The intake process averaged 50 minutes per person, and many citizens with urgent situations — such as essential service cutoffs or domestic violence — could not be seen during office hours.
Previous Approach: In-person service during office hours (8am–5pm, Monday through Friday). Receptionists conducted extensive interviews, often asking questions that could have been answered through standardized forms. Average wait times were 45 minutes, with a 50% abandonment rate.
Stakes: Citizens facing violence, illegal evictions, or denial of healthcare were left without legal guidance. The justice access gap in the service area widened each year.
The Approach
Tools Used: A custom chatbot built on the GPT-4 API, integrated with the Defensoría's case management system. The chatbot handles initial intake, competency assessment, and documentation collection in Spanish and basic Quechua.
Implementation Strategy: Developed over eight months in partnership with a local law school and a technology NGO. The chatbot guides the citizen through a structured intake process available 24/7, determines whether the inquiry falls within the Defensoría's mandate, identifies the category of the affected right, and assesses urgency. High-urgency cases (gender-based violence, risk to life) are automatically escalated. The chatbot does not provide legal advice — it performs triage and prepares cases for attorney review.
Investment: $60,000 USD in initial development (co-funded with international cooperation grants), $15,000 USD/year in hosting and API costs. One legal assistant was retrained as a 'digital triage manager' to monitor and improve the system.
The Results
Quantified Outcomes
- Inquiry capacity increased from 80 to 260 citizens per month (3.25x increase)
- Average intake time reduced from 50 minutes to 15 minutes of staff time per case
- 24/7 availability captured 35% of new inquiries outside office hours
- Inquiry abandonment rate fell from 50% to 12%
- Emergency cases (gender-based violence, risk to life) are now identified and escalated within 8 minutes of first contact
Qualitative Outcomes
- Quechua-speaking citizens reported feeling more comfortable with the bilingual chatbot interface than navigating Spanish-only forms
- Staff attorneys now dedicate their intake time to substantive case evaluation rather than data collection
- University interns receive better-prepared case files, increasing their productivity and learning
The Lessons
What Worked
- Extensive testing with real citizens during the development phase ensured the language was accessible and non-intimidating
- Building a 'speak with a person' option at every step maintained citizen trust
- Automatic urgency detection for gender-based violence and risk-to-life cases prevented dangerous delays
What Didn't
- The chatbot initially struggled with citizens presenting multiple interconnected problems (e.g., eviction + custody + domestic violence)
- Some older citizens felt uncomfortable with the digital interface — the office maintained an in-person option for these cases
Advice
AI triage is one of the highest-impact applications for access to justice. But design the tool with your users, not for your users. Every interaction must feel like a helping hand, not a bureaucratic barrier.
Our Takes
This may be the most meaningful application of AI in legal practice: expanding access to justice for people who would otherwise go unrepresented. The 3x increase in capacity while reducing the abandonment rate from 50% to 12% demonstrates that AI can address the justice gap without compromising the human elements that vulnerable populations need.Lawra (The Moderate)
The numbers are compelling, but I worry about the populations being served. Low-income residents facing eviction or domestic violence are among the most vulnerable legal clients. A chatbot — however well designed — cannot read emotional cues or detect when someone is in immediate danger but downplaying their situation.Lawrena (The Skeptic)
From 80 to 260 citizens per month — that is 180 additional families receiving legal help every month who would have been turned away. And the 24/7 availability capturing 35% of new inquiries outside office hours means people in crisis don't have to wait until Monday. Technology serving humanity at its best.Lawrelai (The Enthusiast)
This is what I mean when I say the real power of technology is to create conditions where people can transcend their circumstances. AI didn't provide legal advice — it removed the bureaucratic barriers preventing people from reaching the human expertise they needed. That is the right paradigm: AI as an enabler of human connection.Carlos Miranda Levy (The Curator)
Sources & References
Have a Success Story to Share?
We're always looking for well-documented examples of AI adoption in legal practice. If your organization has a story worth telling, we'd love to hear from you.
Ready for structured learning? Explore the Learning Program →
Comments
Loading comments...