A technology transfer package built from your development records, with every equipment gap found before the engineering run
Every transfer package built in days, each gap found before a batch.
See one case, screen by screen ↓The set of documents a sending site hands to a receiving plant or CDMO so it can make the same product: the process description, every parameter with its ranges and class, materials, equipment, analytical methods, risk assessments, holds and the open gaps. Technology Transfer Package drafts it from your development records in the company template, with every statement cited.
How to choose AI solutions for life sciences →
Updated
To move a process from development to a commercial plant, or from a plant to a CDMO, the sending site has to hand over everything the receiving site needs to make the same product: the process description, every parameter with its set point, operating and proven ranges and its class, the materials, the equipment, the analytical methods, the FMEA, the hold times and the open gaps. An MS&T lead assembles that package by hand from dozens of development reports, batch records and spreadsheets, and the knowledge that is not written down stays with the people who ran the process.
The most expensive gaps are the ones nobody compared on paper: a receiving-site machine that cannot hold a range the process was validated in. A disc-stack centrifuge that tops out below the validated g-force, a bioreactor with a different sparger, no port for the in-line Raman probe the control strategy relies on — these are too often found in the engineering run, when a batch is already at stake. And when the batch record says one hold time and a newer study says another, someone has to notice before the package goes out.
Estimated split from kick-off to a complete package draft for one biologics drug-substance transfer, by hand and with the solution.
Six agents read the records, compare every step with the receiving site's equipment and write each gap; the transfer lead decides each one.
TT-2611 moves Veltrimab drug substance from the 200 L single-use pilot suite at Process Development — Cambridge, MA to the 2,000 L single-use suite at Plant 3 — Leiden. The protocol is due Oct 14 and engineering run 1 is set for Nov 16. On Oct 5 Leiden sent a new revision of its equipment list. Here is what happens next, screen by screen, in the working solution.
Ten transfers, from development, a plant or an acquired company to a receiving plant or CDMO. TT-2611 is in Protocol at 81% readiness with 12 open gaps, 6 of them to decide, and engineering run 1 is 40 days away. Beside it: the Incheon second source in knowledge capture at 42%, Corventa at Tavora Pharma Services in Puerto Rico at 93% with PPQ batch 1 on Oct 27.
The home screen is the process itself: 11 unit operations, upstream and downstream, each with its critical parameters and the range it was validated in at Cambridge — pH 6.90–7.10, power per volume 15–25 W/m³, a harvest at 10,000–12,000 × g finished within 4 h of the end of culture. The ranges come from 43 development records, 212 parameters in all.
“Pick a receiving site to compare its equipment with every validated range.”
Against EQ-LDN-0207 rev 2, the 11 steps come back 7 fit, 1 adjust, 3 gaps, 1 new. The seed train needs an adjusted setting. The production bioreactor is red with its known gaps — a drilled-hole sparger and no Raman port. Viral filtration is red for a filter from a second supplier. Harvest and clarification turns red, marked new.
Leiden's disc-stack centrifuge DSC-LDN-02 is now limited to 8,000 × g. Matching the validated Q/Σ limits the feed to 410 L/h, so 2,000 L takes 4.9 h against a 4 h window; keeping 4 h means 560 L/h and a Q/Σ 37% above the validated value. The analyst drafts FM-41 for the FMEA draft: severity 7, occurrence 5, detection 4 — risk priority 140, against an action threshold of 100.
Proposed action: “Four scale-down runs at the 8,000 × g equivalent Q/Σ … and 40% more secondary depth-filter area, before the Nov 16 engineering run. Owner Sam Patel, due Oct 30.”
One click adds the line to the FMEA draft and raises G-2611-15: harvest centrifuge limited to 8,000 × g, harvest validated at 10,000–12,000 × g. It is High, blocks the engineering run, and carries the evidence, the proposed action and its owner, Sam Patel, due Oct 30. The FMEA had assessed harvest at the validated g-force only.
The drafted risk line stays a draft until Quality approves the FMEA revision.
Since the last check on Oct 2, three documents arrived: the DR-VLT-0451 hold-time study, Leiden's equipment list rev 2 and the BHT-VLT-04 buffer hold-time study. The Document reader, Parameter extractor, Equipment fit analyst, Checklist checker and Gap writer each run in view, between 1.5 and 4.3 seconds a step. The biologics checklist now reads 30 found, 8 partial, 5 missing, 1 where two sources disagree and 2 not applicable — 77% ready.
“1 new gap written for you to decide. G-2611-03 · Neutralised pool hold: 48 h in the batch record, 72 h in the hold-time study.”
The master batch record MBR-VLT-DS-200 v5.2 holds the neutralised viral-inactivation pool at 2–8 °C for not more than 48 h. The new study DR-VLT-0451 finds it stable for 72 h, with HMW up 0.1% and charge variants unchanged; at 96 h HMW rose 0.4%. Lena records the recommended choice: 48 h as the operating limit, 72 h as the proven limit, with her reason.
“The Parameter extractor found two values for this limit. Agents never pick one — you decide which transfers.”
The receiving lead sees the same comparison from both equipment lists: 11 aspects, 4 fit, 3 adjust, 4 gaps. Matching 20 W/m³ needs a tip speed of about 1.5 m/s on Leiden's SUB-2000, near its 1.6 m/s maximum and above the 0.8–1.3 m/s studied. The drilled-hole sparger is likely to raise pCO₂, there is no Raman port, and the virus filter's clearance claim does not carry over to supplier B.
The signature gate lists every gap that blocks the engineering run. Lena reviews and agrees the five proposed actions still open, each with its owner and due date, and signs the protocol with the meaning “Approved — transfer lead”. Dr. Maya Chen signed for the sending site on Oct 5; Sam Patel signs for Leiden and Omar Haddad approves for Quality. TT-2611 moves to Engineering run.
“Every gap that blocks the engineering run needs an agreed action before anyone signs.”
The tech transfer package for TT-2611 is draft v0.4: purpose and scope, product and process overview, process description, parameters and control strategy, materials, equipment fit and scale-up, analytical methods, risk assessment, holds and stability, and the open gaps with their agreed actions. It can be rebuilt from the latest process knowledge and exported.
The same transfer, seen by the five people who carry it — what their weeks looked like, and what they look like now.
Six agents read the records, compare every step with the receiving site's equipment and write each gap; the transfer lead decides each one.
Reads development reports, batch records, FMEAs, methods and equipment lists; records each one's code, version and effective date, and splits it into passages by unit operation so every value can be cited.
Extracts every process parameter — set point, operating and proven ranges, class (CPP, KPP, non-critical) and scale-dependence — with its source passage, and normalises the units.
Compares the receiving site's equipment with every validated range on the process flow, computes the scale factors — power per volume, tip speed, kLa, Q/Σ — and drafts the FMEA line for a step that cannot hold its range.
Checks the package against the transfer checklist for the modality — 46 items in 10 areas for biologic drug substance — and marks each item found, partial, missing or in conflict.
Writes one gap per missing, partial or conflicting item, with the evidence, severity, what it blocks — the engineering run or PPQ — a proposed action and a suggested owner.
Drafts the tech transfer package, the transfer protocol and the engineering-run sampling plan in the company template, section by section from verified knowledge.
Decides each gap. The agents propose; a named person decides.
Ask in plain words — or say what to draft or change. Every answer points to the passage it came from.
Can Leiden’s centrifuge run the Veltrimab harvest?
Not inside the validated range. The harvest was validated at 10,000–12,000 × g, Q/Σ at most 1.4 × 10⁻⁸ m/s, finished within 4 h. Leiden’s rev 2 equipment list limits DSC-LDN-02 to 8,000 × g: matching Q/Σ means 410 L/h and a 4.9 h harvest; keeping 4 h puts Q/Σ 37% high. Gap G-2611-15 holds it — owner Sam Patel, due Oct 30.
Which hold time should Leiden use for the neutralised pool?
The executed batch record allows 48 h at 2–8 °C; the new hold-time study supports 72 h with HMW up only 0.1%. I recommend transferring 48 h as the operating limit and 72 h as the proven limit, and updating the batch record at its next revision.
Why is pCO₂ a gap at Leiden?
At 200 L peak pCO₂ stayed between 62 and 95 mmHg and was never varied. Leiden’s 2,000 L bioreactor has a 1 mm drilled-hole sparger and overlay capped at 0.1 vvm, so CO₂ will build up more than in Cambridge’s open-pipe design. Proposed: six 3 L runs at 120–180 mmHg before Nov 16.
Draft the engineering-run sampling plan for TT-2611
Drafted — 42 in-process samples across 9 unit operations, built from the parameter table: 18 in the production bioreactor (daily glucose without Raman, pCO₂, viability), 6 at Protein A, 4 at viral inactivation and 14 from polishing to UF/DF. Off-line glucose twice daily replaces Raman. Attached to gap G-2611-12 as evidence.
13 screens from the working solution, on its sample data. Pick one to see it large.
The checklists, the equipment-fit rule, the risk threshold and the gates are settings — changes apply to the next gap check and are on the audit trail.
| Setting | Default | Choose from |
|---|---|---|
| Checklist for biologic drug substance | Biologics drug substance v4 · 46 items, 10 areas | Biologics drug substance v4 · Biologics drug substance v3 |
| A step fits when | the whole validated range | the whole validated range · the normal operating range only |
| Draft a risk line when | a step cannot hold its range | a step cannot hold its range · any step needs a new setting |
| Risk priority that needs an action | above 100 | above 80 · above 100 · above 120 |
| Protocol signature needs an agreed action on every gap that blocks the engineering run | On | On · Off |
| Closing a gap needs both sites | On | On · Off |
| Every scale-dependent parameter needs a scale rationale | Off | On · Off — when on, a missing study or calculation becomes a gap |
| Two sources that disagree always go to a person | Always on | Cannot be turned off — the agents never pick a value |
Every figure is labelled: a target the solution is built to, an estimate, a typical published result, or a proven one.
“demo” = seen in the working solution, on its sample transfer data · “target” = the design goal, measured in the live solution · “estimated” = our estimate · “typical” = published figures (Thermo Fisher Patheon white paper — a typical transfer takes six to nine months, about four months even with a complete package; Bioprocess Online, 2025 — analytical method transfer can take up to 18 months). Context: ICH Q10 (knowledge management) · WHO TRS 1044 Annex 4 (technology transfer, 2022) · ISPE Good Practice Guide: Technology Transfer. People, companies, products and sites named on this page are fictional — characters and sample data in the working solution.
The set of documents a sending site hands to a receiving plant or CDMO so it can make the same product: the process description, every parameter with its ranges and class, materials, equipment, analytical methods, risk assessments, holds and the open gaps. Technology Transfer Package drafts it from your development records in the company template, with every statement cited.
When a receiving site is picked on the process flow, the Equipment fit analyst compares that site's equipment list with every range the process was validated in, step by step, and computes scale factors such as power per volume, tip speed, kLa and Q/Σ. A step the equipment cannot hold turns red, with its FMEA line drafted and a proposed action.
The Checklist checker runs the transfer checklist for the modality — 46 items in 10 areas for biologic drug substance — and marks each item found, partial, missing or in conflict, with cited evidence. The Gap writer turns each finding into a gap with its evidence, severity, what it blocks, a proposed action and an owner.
The Parameter extractor records both values and raises a conflict. The agents never average or pick; a person decides which value transfers and records the reason, as with a pool hold of 48 h in the batch record and 72 h in a new study.
Yes. A CDMO is a receiving site like a plant: its equipment list is compared step by step, and an incomplete list shows as detail missing. Through the CDMO portal each CDMO sees only its own transfer.
Yes. The agents propose and never decide or close a gap. The transfer lead decides each one, a drafted FMEA line stays a draft until Quality approves the revision, the protocol cannot be signed until every blocking gap has an agreed action, and closing a gap needs both sites.
Yes. The checklists by modality, when a step counts as fitting, when a risk line is drafted, the risk priority that needs an action, the gate rules and who signs for each site are settings, and changes apply to the next gap check.
The Agentic Solution Engine builds and deploys it from your requirements and documents — your transfer procedure, checklists, package template and a set of past development records — and it goes live once every quality gate has passed. We will walk you through it on one of your own transfers first.
Sourced answers from papers, patents and reports in seconds, our own disagreements first.
1-min demoCompetitive Pipeline TrackingEvery rival trial, readout and approval in a sourced brief each Monday
1-min demoPreclinical Study Report AutomationA draft GLP report in hours, every number traced to raw data.
1-min demoWe'll run Technology Transfer Package on your development records and a receiving site's equipment list.