Why the cooling ramp is still the recipe
Fixed profiles persist even in plants with PAT installed. What changes when supersaturation becomes a controlled variable instead of an observed one.
ReadBlog
Notes from the hardest step in making medicine.
Crystallization, PAT, autonomy, and the regulatory reality of putting AI in control of a drug-substance batch.
Latest
Recent writing
New posts on control, sensing, and quality.
Four instruments, one crystallization state
Raman, FBRM, PVM, and NIR each under-determine the slurry. Fusion is not a nice-to-have — it is the only way to get a decision.
ReadPolymorphs do not care about your schedule
Form transitions during drying are the failure nobody plans for. Predicting them in-line changes what isolation looks like.
ReadCrystallization
On the crystal itself
What QA actually asks an AI vendor
The six questions that end most industrial-AI pitches in pharma, and how to build a product that survives them.
ReadTech transfer as a prediction problem
A twin that maps lab crystallization to plant behaviour compresses the slowest step in bringing a molecule to scale.
ReadBounded autonomy in a GMP plant
Shadow, assist, and auto-control are not marketing tiers — they are a validation strategy.
ReadPlant & platform
On running it in production
Where the yield actually goes
Conservative profiles protect quality by leaving product in the mother liquor. Live feedback buys back the margin.
ReadEdge-first, because batches do not pause
Why perception and control inference belong at the train and training belongs in the cloud.
ReadEvery correction is training data
The chemist who overrides an agent is the most valuable data source in the plant. Capturing that is the moat.
ReadEditorial position
We write about what we can defend
Polymorra is pre-commercial. Where a claim is a design target rather than a measured result, we mark it [ASPIRATIONAL]. Where a figure will come from design-partner campaigns, we mark it [PLACEHOLDER]. We would rather be trusted than impressive.
- Design targets are labelled as targets
- Market figures are labelled as internal estimates
- Customer outcomes wait until customers publish them
The autonomous loop
Perceive, plan, act, sense, optimize, log
Polymorra closes the loop around the reaction mass and the crystal — the two things a conventional DCS cannot actually see.
Perceive
Fuse Raman, FBRM, PVM in-situ imaging, NIR, and reactor telemetry into a live picture of the reaction mass, slurry, and solids.
Plan
Plan the react-and-charge, crystallize-and-seed, and isolate-and-dry moves for this batch against the target polymorph, particle size, and purity.
Act
Run with adaptive control: cooling and antisolvent profile, seeding, supersaturation, agitation, addition rate, and endpoint.
Sense & predict
Predict polymorph, particle size and habit, impurity profile, and yield in-line — before offline XRPD and HPLC confirm it.
Optimize
Optimize yield, solvent use, cycle time, and reprocessing risk across the campaign, then flag off-spec and polymorph risk early.
Log & retrain
Write an immutable Part 11 record; every chemist correction trains the site model and compounds the data moat.
Plant-edge console
One screen for the batch that is running right now
Supersaturation, particle size distribution, predicted polymorph, impurity trajectory, and the next control move — with the evidence behind each number.
Polymorra plant-edge console showing live supersaturation, particle size distribution, predicted polymorphic form, and impurity trajectory for a running crystallization, with the recommended next control move.
Integrations
Connects to the systems already running your plant
Polymorra reads and writes through the reactor DCS, PAT instruments, crystallizer and isolation skids, historians, and MES you already validated.
Reactor control & DCS
Emerson DeltaVSiemens PCS 7Honeywell ExperionOPC UAModbus/TCP
PAT & analytics
Mettler-Toledo FBRMPVM in-situ imagingReactIREndress+Hauser RamanNIRHPLC / XRPD imports
Isolation & drying
Nutsche filter-dryersCentrifugesAgitated dryersConical millsMicronizers
MES, historian & quality
Körber PAS-XEmerson SyncadeOSIsoft PILIMSeBR / QMS exports
Trust & compliance
Validated for GMP manufacturing
Polymorra is engineered for regulated drug-substance production: grounded outputs, immutable audit trails, graduated autonomy, and validation documentation from day one.
21 CFR Part 11
Immutable, validated audit log for every agent perception, recommendation, and control action, with e-signature-ready review flows.
GMP / ICH Q7
Change control, validated model versioning, and deployment documentation designed for drug-substance manufacturing.
SOC 2
SOC 2 Type I in progress, Type II on the roadmap; SSO/RBAC, encryption in transit and at rest. [ASPIRATIONAL]
Route & IP protection
Per-tenant isolation of routes, recipes, spectra, and crystal images. No cross-tenant training. On-prem option.
Graduated autonomy
Shadow → assist → bounded auto-control. Every autonomy level is explicitly configured, bounded, and revocable.
Grounded outputs
Every recommendation cites the spectra, images, telemetry, SOP, or specification it was derived from.
Plant-edge first
Inference runs on-site so control loops survive network loss; cloud is for training, fleet, and analytics.
Human-in-the-loop
Chemist and QA checkpoints are first-class: approve, correct, or reject — and every correction trains the models.
Voices from the plant
What operators say
Composite quotes from design-partner discovery interviews. [PLACEHOLDER — to be replaced with named references post-pilot]
“The crystallization step is where our campaigns live or die. Seeing supersaturation, particle size, and predicted polymorph in one place — and having the profile adjust itself — is what we have wanted for twenty years.”
“We do not want another dashboard. We want the batch to come out right and the record to be reviewable by exception. That is the only pitch that gets past QA.”
“Tech transfer is our bottleneck. If the twin can predict the plant crystal from the lab crystal, that is worth more than the software costs.”
FAQ
Frequently asked questions
Both — in a deliberate order. Polymorra starts in shadow mode (perceive only), moves to assist (recommend, chemist approves), then to bounded auto-control on qualified trains where setpoint moves stay inside validated control windows. Every level is configured per train and revocable.
Polymorra is built for GMP (ICH Q7) and 21 CFR Part 11: immutable audit trail, validated model version locks with rollback, change control, grounded and citable outputs, and human-in-the-loop checkpoints. We ship a validation package and support IQ/OQ/PQ activities with your quality unit.
API or OPC access to your reactor DCS, PAT instruments (Raman, FBRM, PVM, NIR), crystallizer and isolation skids, and MES or historian — plus a train with enough campaign volume for clear ROI.
No. Per-tenant isolation of routes, recipes, spectra, and crystal images is enforced at the data and model layer. Cross-site benchmarks are opt-in and de-identified.
A typical wedge pilot instruments one train, runs shadow mode over a baseline campaign, then moves to assist mode with a defined polymorph, yield, or reprocessing success metric agreed up front.
Perception and control inference run at the plant edge. Cloud handles training, fleet management, and analytics — losing it degrades reporting, not the batch.
Want us to write about your problem?
Start with one crystallizer or reactor train. Prove polymorph, yield, and reprocessing ROI in a validation-friendly pilot. Then expand across the plant.