Postpartum Psychiatric Risk Monitoring Agent using LangChain and Agno
Monitoring drug liability and psychiatric risk during postpartum clinical care requires synthesizing clinical records and adverse event data. Developers will build a LangChain and Agno stateful agent to extract medication histories, assess liability risk markers, and generate structured clinical review alerts with complete auditability.
What you are building
The core problem, expected build, and operating context for this challenge.
Develop a stateful clinical agent using LangChain and Agno to evaluate postpartum psychiatric medication histories and risk factors.
How work is evaluated
Evaluates medication extraction recall, polypharmacy risk score calculation, and audit logging validity.
Challenge datasets
Public datasets, plus the private datasets you have access to.
How submissions are scored
These dimensions define what the evaluator checks and which criteria separate a passable run from a strong one.
review_flag_test
Verifies that high-risk polypharmacy triggers clinician review.
This dimension contributes its full weight only when the submission satisfies the requirement. Partial credit is not awarded.
entity_extraction_f1
F1 score of medication name extraction from EHR text. • target: 0.92 • range: 0-1
This dimension contributes its full weight only when the submission satisfies the requirement. Partial credit is not awarded.
What you should walk away with
Configure a LangChain stateful processing graph for electronic health record (EHR) text.
Integrate Agno agent structures for multi-modal drug interaction monitoring.
Extract psychotropic polypharmacy patterns from patient encounter notes.
Generate structured clinical liability and safety alerts with explicit provenance links.
Reference links and supporting material
Synthetic postpartum patient clinical notes containing realistic medication histories and symptoms.
Challenge setup
Evaluates medication extraction recall, polypharmacy risk score calculation, and audit logging validity.
Challenge input
JSON containing clinical encounter note text.
LangChain
Provides structured chain execution and state management.
Agno
High-performance agent library for domain-specific tasks.
Evaluated output
JSON containing extracted medications, risk level, and review alert flag.
- Verifies that high-risk polypharmacy triggers clinician review.
- F1 score of medication name extraction from EHR text. • target: 0.92 • range: 0-1
- Benchmark: Clinical Adverse Event Detection Benchmark
- Entity Extraction F1 target: 0.92
- 1 public reference case
- Python execution harness
- Python sandbox (unavailable on Versalist)
Summary shows the first task, up to three tools, and up to three evidence items. Open the technical recipe to view all configured items.
View technical recipe
Configured tools
- Langchain · Required
- Agno · Optional
- PyTorch Lightning · Optional
- Langchain · Required
- Agno · Optional
Evaluation contract
- review_flag_test · Weight 1
- entity_extraction_f1 · Weight 1
Metrics
- entity_extraction_f1 · Target 0.92
Benchmarks
- Clinical Adverse Event Detection Benchmark
- biotech
Reference cases
- 1 public reference case
Workflow tasks
- Task 1
Input: JSON containing clinical encounter note text.
Output: JSON containing extracted medications, risk level, and review alert flag.
Execution harness: Python
Evaluation tools
- LangChain · LLM Orchestration Framework
- Agno · Agent Intelligence System
Recipe freeze marker
No recipe freeze time is recorded.