Life Saving Agent

Life-Saving Agent connects hospitals to blood centers in real time, dispatches blood fast, schedules donor visits smartly, and answers organ/patient queries(future)—saving lives with speed and care.

Comment 7

Inspiration In an emergency room, a doctor runs out of time—and blood. Somewhere nearby, a blood center has exactly what’s needed. But without the right system, that connection may come too late. We were inspired by the need to eliminate this gap, using technology to save lives faster and smarter.

What it does Life-Saving Agent is an AI-powered assistant built on Agentforce that connects hospitals, blood centers, donors, and patients in real time. It instantly matches hospital blood requests with nearby blood centers based on blood type, units, and location. If an exact match isn’t available, it finds the best alternative and assigns a delivery agent.

It registers donors, detects their intent, collects prerequisites and location, and schedules future appointments at nearby centers. It also answers organ donation questions, patient inquiries, and books future appointments (extension)—through natural, conversational interaction.
How we built it Front-end: Blood Request from Hospital

  1. Accounts are used to store blood centers and hospitals, differentiated by fields and record type.
  2. Blood Inventory is a child of Account (hospital & blood center) that stores blood type, units available, and last updated date.
  3. Blood Request is a child of Account (hospital) that stores request information submitted by the hospital. It acts as a junction object between Blood Inventory and the hospital.
  4. The standard Case object is used for follow-up.

Register Donor

  1. Standard Contact is used to store donor information.
  2. Standard FSL Service Appointment is used to schedule appointments for contacts (donors).
  3. Service Territories are used to store blood center details, with a lookup to Account.
  4. Operating Hours are used to set the available hours of blood centers for appointments.
  5. Experience Cloud is used to provide a user-facing portal for customers.

Back-end

  1. Flow is used to create a blood request. Apex code matches the distance between the hospital and blood center.
  2. Another flow registers a donor, and Apex code schedules an appointment by checking operating hours.
  3. Database: Salesforce for now (API requests to blood centers are planned for the future).
  4. AI/NLP: Intent recognition and handling of alternative blood types.
  5. Maps & Location Matching: Implemented using Apex code.

Challenges we ran into

  1. Training intent recognition to handle natural human queries for organ and blood donation.
  2. Coordinating multiple roles—hospitals, donors, blood centers—in the object data model.
  3. Handling edge cases like no immediate blood match and implementing fallback delivery logic.

Accomplishments that we're proud of

  1. Successfully built a working prototype that can match hospital requests with blood centers and schedule donors.
  2. Implemented a smart fallback system for unavailable blood types with delivery support.
  3. Created an interactive AI assistant capable of handling real-world donation and inquiry scenarios.

What we learned

  1. How to build and scale real-time matching systems with multiple variables.
  2. Integrating AI-driven intent recognition into healthcare use cases.
  3. The importance of Agentforce.
  4. Working under time pressure to turn an idea into a solution.

Built With

+ 20 more
Share this project:

Updates