Scientific Information Retrieval System™
Retrieval built for how scientists ask questions
SIRS reads abbreviations, asset codes and drug classes the way your team does, so the evidence it returns matches the question that was asked.
How SIRS works
How does SIRS work?
SIRS runs on two knowledge graphs, the Bio Graph and the Pipeline Graph, and handles a scientific question in five steps, from reading what is being asked to grounding every claim in its source. Standard retrieval pipelines rely on similarity search alone, while SIRS expands the question with ontologies, searches by keyword, meaning and knowledge graph, reads tables and figures, and ranks evidence by mechanism before any answer is written, with each step calibrated by biomedical information specialists.
Understand intent
SIRS reads what the question is asking for. It reformulates ambiguous or abbreviated queries, extracts the key biomedical entities and recognizes the type of answer needed, such as a safety profile, a mechanism or a trial history.
Expand with ontologies
Curated ontologies resolve drug classes, gene synonyms and disease names to the concepts behind them, so a class such as anti-CD20 monoclonal antibodies reaches all 28 drugs in that class.
Retrieve the evidence
Keyword, meaning and knowledge graph search run together across literature and pipeline, so evidence is found whether it matches by exact term, by meaning or by biological relationship. SIRS also reads tables and figures inside each document. Answers can also show relevant figures from papers, including papers that are not cited.
Rank by mechanism
Because the Bio Graph records the direction of every relationship, SIRS can rank evidence by its effect. Findings that point the right way come first, while opposite or off-topic findings move down the list.
Ground every claim
SIRS extracts insights from each document, grounds them in the source, labels each one as a direct answer, related evidence or no answer, then writes the summary. Every statement links to a cited excerpt and unsupported claims are dropped.
FXR activation represses CYP7A1 1 and reduces conversion of cholesterol to bile acids 2, a mechanism linked to raised LDL-C in FXR agonist trials 3.
Two knowledge graphs
What do the knowledge graphs add to retrieval?
Text search finds passages that mention the right words. The two Causaly knowledge graphs add structure that text alone does not carry, the direction of a biological effect and the state of every development programme, so SIRS can retrieve and rank by what the evidence actually says.
Cause and effect, with direction
Directional cause and effect across targets, drugs and diseases, extracted from the literature and linked back to the source sentence. Every relationship carries its direction, and refuting statements are kept as well as supporting ones.
Who is developing what, and at which stage
Development programmes across the industry, showing who is developing what and at which stage. SIRS uses it to add programme history, discontinuations and competitor activity to answers that would otherwise rely on whatever web search turns up.
How does SIRS keep unsupported claims out of the answer?
Every insight is checked against the excerpt it came from before synthesis, and each statement in the final answer links to the highlighted passage in its source, so reviewers can verify it in seconds.
Extract insights
Relevant insights, not raw chunks, are taken from each document.
Ground and validate
Each insight is mapped to a cited excerpt, and unsupported ones are removed.
Classify the evidence
Direct answer, related evidence or no answer, which keeps limits visible.
Synthesize with citations
Supported insights form one answer with sentence-level citations.
Clinical program evaluation
“If it outputs a table, I can pull the exact references behind each point. It’s faster and reduces the risk of missing key perspectives.”
What evidence does SIRS search?
SIRS searches the public biomedical record through one pipeline, and your own scientific documents too once you choose to connect them. It starts with the sources below, then reaches further through live search or specialist connectors when a question needs it. New publications are searchable within 24 to 48 hours, and the Bio Graph is refreshed every two to four weeks.
The same retrieval over your internal documents
Investigator brochures, study reports and toxicology reports are parsed down to the section, table, row and figure in a dedicated single-tenant environment.
Frequently asked questions about SIRS
What is the Scientific Information Retrieval System (SIRS)?
SIRS is the retrieval layer of the Causaly platform. It interprets a scientific question, expands it with ontologies, retrieves evidence from public and internal sources through keyword, meaning and knowledge graph search across the Bio Graph and Pipeline Graph, reads tables and figures, ranks that evidence by mechanism, and grounds every claim before it reaches Causaly’s agents.
How is SIRS different from standard RAG or vector search?
Standard RAG pipelines rely on vector similarity alone and return a short list of loosely related passages. SIRS adds exact-term matching for asset codes and study IDs, ontology expansion for synonyms and drug classes, knowledge graph retrieval for mechanisms and pipeline programmes, and table and figure retrieval, then ranks the combined results by direction of effect before any answer is written.
How does SIRS reduce hallucinations?
SIRS extracts insights from each document, grounds each one in a cited excerpt from its source, labels it as a direct answer, related evidence or no answer, then writes the summary. Claims without support are dropped, so the response states its own limits.
Can SIRS search our internal documents?
Yes, on request. Private Data deployments run SIRS over internal documents such as investigator brochures, clinical study reports and toxicology reports in a dedicated single-tenant environment, retrieving down to the section, table, row and figure with provenance preserved.
How current is the evidence SIRS searches?
New publications are searchable within 24 to 48 hours of publication and are added to the Bio Graph within two to four weeks.
Does Causaly train models on customer data?
No. Customer data is not used to train or fine-tune models, and model providers operate under zero-data-retention terms.
Go deeper into how SIRS retrieves and grounds scientific evidence
The whitepaper covers the full retrieval architecture and benchmark method, and Causaly lets you see SIRS-powered answers on your own research questions.
