Healthcare AI · Built on AWS

Clinical knowledge,
instantly retrieved

CliniqAI is an AI-powered assistant that lets healthcare professionals ask clinical questions in plain language and receive grounded, cited answers from institutional guidelines and formularies — in seconds.

View the codebase See how it works

CliniqAI — query assistant

"What is the recommended first-line treatment for Type 2 diabetes with CKD stage 3?"

Answer

For patients with T2DM and CKD stage 3 (eGFR 30–59), metformin may be continued if eGFR ≥30 with dose reduction. SGLT2 inhibitors (empagliflozin, dapagliflozin) are recommended as first-line add-on therapy due to demonstrated renoprotective effects. GLP-1 receptor agonists are preferred over sulfonylureas for their lower hypoglycaemia risk in this population.

ADA Standards of Care 2024 · p.48 KDIGO CKD Guidelines · p.112 Hospital Formulary v4 · p.23

Clinicians spend too much time finding answers

Healthcare professionals navigate dozens of fragmented documents — PDFs, intranets, EHR portals — to find evidence they need at the point of care. CliniqAI consolidates institutional knowledge into one query interface.

23min

Average time physicians spend on administrative and information lookup tasks per patient encounter

40%

Of clinical decisions are made without consulting current guidelines due to time constraints

6,000+

Drugs in a typical hospital formulary — impossible to memorise dosing, interactions, and adjustments

How CliniqAI works

A serverless RAG pipeline on AWS — from question to cited clinical answer in under 3 seconds

Authenticate

Clinician signs in via Cognito with MFA. Role-based access scopes which documents are searchable.

Ask

Natural language query is submitted via API Gateway, validated by WAF, and routed to Lambda.

Retrieve

Bedrock embeds the query; OpenSearch Serverless finds the most semantically relevant document chunks.

Generate

Claude 3.5 Sonnet synthesises a grounded answer from retrieved context only. No hallucination.

Cite

Every answer includes source document, section, and page number so clinicians can verify the reference.

AWS architecture

20+ AWS services across security, AI, compute, and observability layers

Security Identity, encryption & threat protection

Amazon Cognito AWS WAF AWS KMS AWS Secrets Manager IAM Amazon GuardDuty

Networking Private network boundary

Amazon VPC VPC Endpoints (Bedrock, S3, OpenSearch)

API Request routing & authorisation

Amazon API Gateway

Compute RAG orchestration & ingestion

AWS Lambda (query handler) AWS Lambda (ingestion handler) Amazon EventBridge

AI / Search Foundation models, embeddings & vector search

Amazon Bedrock (Claude 3.5 Sonnet) Bedrock Knowledge Base Bedrock Guardrails Amazon Titan Embeddings V2 Amazon OpenSearch Serverless

Storage Document store & audit logs

Amazon S3 (clinical documents) Amazon S3 (CloudTrail logs)

Observability Monitoring, compliance & audit

Amazon CloudWatch AWS X-Ray AWS CloudTrail AWS Config Amazon SNS

Built for HIPAA compliance

CliniqAI is designed around HIPAA Technical Safeguard requirements from the ground up

Access control

Cognito user pools with enforced MFA, role-based document scoping, IAM least-privilege roles, and VPC network isolation

Audit controls

CloudTrail logs every API call with timestamp and caller identity. CloudWatch structured logs. X-Ray distributed tracing.

Data integrity

KMS encryption at rest for S3 and OpenSearch. TLS 1.2+ in transit. S3 versioning prevents accidental data loss.

Transmission security

VPC endpoints route all traffic between Lambda, Bedrock, S3, and OpenSearch privately — never over the public internet.

Automatic logoff

Cognito access tokens expire after 60 minutes. Refresh tokens are revocable immediately if a device is compromised.

Threat detection

GuardDuty monitors for anomalous access. WAF blocks injection attacks, bots, and malicious IPs before they reach the API.

About CliniqAI

CliniqAI is an early-stage healthcare AI startup building tools that reduce the time clinicians spend looking for information — so they can spend more time with patients.

Our first product is a HIPAA-aligned RAG assistant that connects to a hospital's existing document library and lets any authorised clinician ask questions in plain language, receiving grounded answers with citations in under three seconds.

We are a bootstrapped startup building on Amazon Web Services, taking advantage of the AWS serverless and AI ecosystem to deliver production-grade infrastructure without the overhead of managing servers.

The full infrastructure is open-source and available on GitHub, built with Terraform, Python, and Amazon Bedrock.

View source on GitHub

Company CliniqAI

Founded 2025

Stage Pre-seed / Bootstrapped

Primary cloud Amazon Web Services

Sector Healthcare AI / Clinical informatics

Target customers Hospitals, clinics, health systems

Infrastructure Fully serverless, open-source

Clinical knowledge,instantly retrieved