Hospitals and healthcare facilities face increasing patient loads and a declining number of clinical staff to care for them. As a result, these organizations are desperate for ways to meet patient care demands with shrinking resources.
One key way to confront this challenge is to reduce administrative tasks and streamline clinical workflows and communication. Doctors and nurses need fast ways to get correct, reliable information.
MediPulse had a vision for an AI-powered solution to meet this growing, critical need. They wanted to empower hospital staff to ask natural-language questions against a large and diverse body of organizational policies, standard operating procedures, and training materials.
These questions are unpredictable in both scope and complexity. For example, delivering the correct answer to “What is the protocol for patient discharge after cardiac surgery?” requires finding and synthesizing information across multiple documents. “Can a nurse practitioner authorize this medication during the night shift?” requires the system to understand role-based policies, shift-specific rules, and drug-specific protocols simultaneously.
A retrieval pipeline with fixed logic cannot reliably handle these types of queries. To provide trustworthy answers, the system needs to reason about which documents to retrieve, evaluate whether the retrieved context adequately answers the question, and decide whether to search again with reformulated parameters.
To bring their vision to life, MediPulse needed an AI solution that could evaluate the quality of its own intermediate results and adapt its behavior accordingly.
MediPulse partnered with Taazaa for the skills and experience needed to build this advanced healthcare AI product.
To achieve all of MediPulse’s goals, Taazaa built a LangGraph ReAct agent—the Reasoning and Acting pattern where the AI observes the result of each step and decides its next action.
The agent selects from available tools and retrieves documents using a three-level hierarchical RAG system (global policies, tenant-specific policies, and site-specific policies), with strict priority resolution that ensures a site-level policy always overrides a global default.
The retrieval mechanism itself is two-stage. Utilizing AWS Bedrock Titan embeddings and pgvector, a hybrid vector-plus-keyword search retrieves the top 50 candidate passages. A cross-encoder reranker then narrows the results to the top two most relevant passages.
This two-stage approach eliminates a well-known failure mode: pure vector search can return passages that are semantically similar to the query but factually incorrect for the specific context.
What makes the system genuinely agentic is the decision loop. The agent does not follow a predetermined path. It evaluates retrieval quality, reformulates queries when results are insufficient, and maintains per-user semantic memory via Langmem backed by pgvector.
In practical terms, it means that the AI’s response to a follow-up question correctly inherits the context of the original query. For example, asking, “What is the protocol for patient discharge after cardiac surgery?” and following it up with, “What about for pediatric patients?” correctly inherits the cardiac discharge context without the user having to restate it. Input classification runs asynchronously in parallel for analytics, ensuring the reasoning loop is not bottlenecked by secondary processing.
This is a prime example of an agentic architecture. The problem space is open-ended, the retrieval targets are heterogeneous, the quality of any single retrieval step is unpredictable, and the system must reason about adequacy before responding. A deterministic pipeline would either return incorrect answers to complex queries or require so many hardcoded branches that it would become unmaintainable.
The HIPAA-compliant AI solution is designed to help healthcare organizations centralize operational knowledge, standardize processes, improve staff engagement, and reduce administrative overhead through a unified digital platform.
MediPulse expects the product to help healthcare organizations improve compliance with organizational policies and training requirements by reducing the reliance on manual, paper-based processes and increasing visibility into staff engagement and training.
The solution helps disseminate information faster across teams and departments, and improves process standardization and consistency across sites. With centralized tracking and reporting, healthcare organizations will be better prepared for audits.
By giving healthcare staff a single, easy-to-use platform for learning, communication, and trustworthy resources, MediPulse helps healthcare organizations deliver quality care even with limited resources.
It’s another project that underscores Taazaa’s commitment to improving people’s lives through great digital solutions.
By giving healthcare staff a single, easy-to-use platform for learning, communication, and trustworthy resources, MediPulse helps healthcare organizations rapidly deliver quality care.
