El Dorado Assist
Prototype
This prototype is designed to demonstrate how El Dorado Assist could help residents in three areas: housing, food, and mental health.
The knowledge base could be expanded to provide support for these additional areas (resources currently available on EDCOE website): Animals, Breastfeeding, Disaster Response, Childcare, Clothing/Diapers, Dental, Employment Education, Family Planning, Food Banks & Services, Legal Services, Medical Offices, Money Assistance, Pharmacy, Utilities, Relationships, Seniors, Substance Use, Support Groups, Telehealth, and Transportation.
A smarter way to connect residents with help — and help community leaders see where the gaps are.
The Opportunity
El Dorado County has many organizations and public agencies providing food, affordable housing, mental-health services and other forms of assistance. Yet having resources is not the same as being able to access them. Residents may not know where to begin, may encounter outdated or fragmented information, or may face barriers involving transportation, hours, eligibility, language, waitlists, documentation, technology, or geography. For Spanish-speaking residents and people already under stress, navigating multiple websites and agencies can be especially difficult.
El Dorado Assist is a locally focused conversational AI navigator designed to reduce that burden. A working prototype currently covers three areas — affordable housing, food, and mental health — using a curated El Dorado County knowledge base. Rather than requiring residents to understand the service system, it allows them to describe their situation in ordinary language and receive guided, personalized help.
Three Ways a Resident Might Seek Help
| Basic Search | Generic AI | El Dorado Assist | |
|---|---|---|---|
| Experience | Returns websites, directories and agency listings. | Understands the situation, can ask questions, summarize options and converse in Spanish. | Combines conversational AI with a curated, locally maintained knowledge base. |
| Resident burden | Resident must sort, compare and verify information. | Much lower, but the AI may depend on incomplete, inconsistent or outdated public information. | Low: the assistant can ask about location, urgency, transportation, language and other barriers before recommending options. |
| Local reliability | Varies by source and search result. | Can be strong, but is not inherently tied to a verified El Dorado resource system. | Designed around locally verified information that can be continually updated. |
| Community learning | Search activity provides little insight into whether people actually found help. | Helps the individual, but community leaders generally do not receive systematic local feedback. | Can aggregate anonymous patterns about needs, barriers, referral outcomes, geographic gaps and missing services. |
SEARCH: “Here are some resources.”
GENERIC AI: “Let me help you find resources.”
EL DORADO ASSIST: “Let me understand what you need and help you navigate the resources available here.”
How El Dorado Assist Would Work
Residents could interact in English or Spanish and simply explain what is happening. The assistant can ask a small number of relevant follow-up questions, search the verified local knowledge base, explain why a resource may fit, and identify practical barriers before sending someone on an unproductive referral.
The knowledge base can include eligibility, service area, hours, languages, transportation or delivery options, appointment requirements, contact information, and the date each listing was last verified.
The knowledge base is a community asset, not merely chatbot content. Participating organizations can help verify their information, and the system can flag records for periodic review. As the project develops, the knowledge base can expand beyond the initial three service areas.
Community Intelligence: Learning From Every Interaction
The distinctive value of El Dorado Assist is not only better navigation. With appropriate privacy protections, anonymous interaction data can create a continuously improving picture of community needs and service performance.
Leaders could see what residents are seeking, where needs occur, why referrals fail, which resources appear to work well, and what residents need for which no adequate resource exists.
What Isn't Working
- 63 residents could not reach an appropriate food resource because of location, transportation, or hours.
- 41 households sought affordable housing for which no currently available option could be identified.
- 27 residents reported difficulty finding Spanish-language mental-health services.
- Rural communities showed a disproportionate share of transportation-related barriers.
What Appears to Be Working
- 82% of residents seeking food assistance were matched with at least one potentially appropriate resource.
- Several programs were repeatedly identified by residents as accessible and responsive.
- Spanish-speaking residents could navigate the same resource system conversationally without searching English-language agency websites.
- Successful referral patterns can identify programs worth strengthening or replicating.
Note: These figures are illustrative, not findings about El Dorado County. They demonstrate the kinds of information a pilot could generate.
A Practical Pilot
The current prototype provides a starting point for a limited pilot focused on affordable housing, food, and mental health. A pilot could bring together county agencies, schools, nonprofit and faith-based providers, healthcare organizations, libraries, philanthropy, and residents.
The immediate goals would be to test whether conversational navigation improves access, establish a reliable process for maintaining the knowledge base, and determine which anonymous analytics are most useful for community planning.
The Case for El Dorado Assist
El Dorado Assist is not primarily an AI project. It is community infrastructure: a way to make an existing network of services easier to navigate and, at the same time, make that network more responsive.
Traditional search tells residents what information is online. Generic AI can help interpret that information. El Dorado Assist adds something neither provides: a continuously maintained local knowledge base combined with a community feedback system.
Help community leaders understand the resources we need.