For deviation, OOS and market complaint investigations
Syntriq works through the evidence behind an investigation, traces every finding to the document it came from, and reviews the result the way an inspector would — against your own procedures and the regulations your site is inspected under.
Works alongside your existing QMS. No system to replace.
The replacement filter was used with a shortened pre-flush, leaving the membrane able to absorb preservative from the first part of the fill12.
Likely
contribution
4
sources
0
quotes against it
1Batch manufacturing record · page 4
Pre-flush volume: 20 L 5 L
2Filter qualification report · §6.2
Minimum pre-flush of 20 L required before product contact.
Batch records, logbooks, lab systems, monitoring data and SOPs live in different systems — some on paper. A QMS holds the report, not the evidence.
Most of an investigation is spent locating records and cross-checking them, leaving the least time for the root cause.
"Human error", a cause no record shows happened, a batch released on a retest: the findings inspectors raise are the ones the record cannot support.
How it works
The stages of a formal investigation, each prepared from the evidence and each decided by the investigator.
Record the event and upload the evidence as it exists: SOPs, batch records including handwritten entries, lab data, equipment logs, sensor exports. Syntriq reads the reference, dates and batches from the documents.
Events from every document are put in order, excursions in monitoring data are found, and the records are checked against your own procedures to flag where they were not followed.
What the evidence cannot yet settle, and which record would settle it — who holds it and what to look for in it.
Which batches are in scope and why, what the evidence shows about product and patient, and where your own classification definitions point. The decisions stay yours.
A fishbone populated with causes drawn from the evidence, 5 Whys and hypothesis testing, and a weighing of each cause. A probable root cause is named only when the evidence clearly supports one.
Actions planned against each root cause with measurable effectiveness criteria, and a justified decision for every batch — ready to go into your QMS.
At any point, ask a question of the evidence and get an answer tied to the documents.
Inspection mode
Review any stage, or the whole investigation, before it leaves your hands. Each finding names the procedure clause or regulation it rests on, says whether that is a requirement or a recommendation, and what would make it hold up.
Operator error named as root cause without the required questions answered
SOP-QA-011 §8.2 · EU GMP Ch.1 1.4(xiv)
Eliminated causes have no evidence recorded for their elimination
SOP-QA-011 §8.1 · 21 CFR 211.192
Procedure sets no maximum number of retests
FDA OOS guidance (2022) IV.B.1 — recommendation
Your procedure requires 3 things here
Answer the procedural, training, task-design and detection questions before recording human error.
SOP-QA-011 §8.2Recorded: Per-cause choice
Confirm or eliminate each contributing factor, recording why it was ruled out.
SOP-QA-011 §8.1Recorded: In the fishbone
Distinguish immediate causes from systemic ones.
SOP-QA-011 §8.3Recorded: Per-cause choice
Your procedure, not a template
Syntriq reads the procedures you upload and applies them where they matter: what your SOP specifically requires of impact assessment, root cause, CAPA and disposition is shown in each stage, and the suggestions are prepared to meet it. Where your procedure says nothing in particular, the standard approach applies.
Regulations
Each requirement is entered from the regulation or guideline itself, with its version, split by the kind of investigation it applies to — so an OOS is never judged against deviation requirements. Choose the frameworks your site is inspected under.
US FDA
21 CFR 211 · OOS guidance (Rev. 1, 2022) · Data integrity Q&A (2018)
EU GMP / PIC/S
EudraLex Vol. 4 Ch. 1, 6, 8 · PIC/S PE 009-17
UK MHRA
OOS & OOT investigations — Phase Ia, Ib, II, III
ICH
Q7 · Q9(R1) · Q10
WHO
GMP main principles, TRS 986 Annex 2
India
Revised Schedule M, G.S.R. 922(E), 2023
Reliability
An investigation tool is only useful if a reviewer can check what it says. These are the rules it works to.
Each finding links to the passage it came from, highlighted in the original document. Anything that cannot be traced to a document is not used.
Where the evidence does not answer a question, the tool says so and names the record that would — it does not fill the gap.
Limit comparisons, durations and breaches are calculated by fixed rules, with units checked, not estimated by a language model.
Syntriq prepares and challenges. The investigator decides, and every change to a prepared reading is kept with its reason.
Your QMS
Changing a QMS means revalidation, migration and retraining. Syntriq asks for none of that. Your QMS stays the system of record, with its approvals and signatures; what Syntriq produces goes into it. Try it on a handful of investigations without changing any existing system.
Where the evidence lives
Batch records · lab systems · equipment and maintenance logs · monitoring data · SOPs · paper
Syntriq
Reads the evidence, prepares each stage, and challenges the result against your procedures and regulations
Your QMS
The system of record: the investigation, approvals, signatures and CAPA tracking
Pilot programme
We are working with a small number of quality teams to shape Syntriq around real investigations. A pilot runs on a few of your own deviations, OOS results or complaints, against your own procedures, with no change to your existing systems.