What changes when a pharmacovigilance leader moves from a mature, legacy-heavy portfolio to a young, fast-growing one — and what does PV leadership actually look like once you strip away the checklists?

In this milestone first-guest episode of What's Next in PV, Tereza Korecká and Jan Kolouch sit down with Pav Rishiraj, Head of Pharmacovigilance at Celltrion and Chair of the ABPI Pharmacovigilance Expert Network. After nearly a decade across big pharma and most recently Ipsen as Director of PV and UK QPPV, Pav shares what actually surprised him in his first 100 days in a very different kind of organization.

The conversation covers:

• Why scaling PV isn't about adding more people and process, and how systems thinking prevents unnecessary complexity
• What actually differs in biosimilar PV versus small molecules and novel therapeutics
• Why pharmacovigilance is "training rich but education poor" — and the real difference between training and competency
• How to build AI governance around explainability, traceability, and human-in-the-loop oversight, with lessons from a real-world system outage
• What "predictive safety" concretely means for case processing and signal management, beyond the buzzword
• What separates a strong PV specialist from someone ready to lead the function
• Where PV is headed in the next five years, and why ten-year predictions are not worth making yet

This is a practical conversation for PV leaders, aspiring leaders, and anyone trying to figure out how much of their team's future work will still be human.

Hosted by Tereza Korecká and Jan Kolouch | Guest: Pav Rishiraj, Head of Pharmacovigilance, Celltrion

Keywords:
pharmacovigilance, PV leadership, biosimilars, predictive safety, AI in pharmacovigilance, AI governance, human in the loop, signal detection, PSUR, competency-based leadership, systems thinking, ABPI, Celltrion

(00:00) - Intro music and episode introduction

(01:02) - What surprised Pav most in his first 100 days at Celltrion

(03:19) - How portfolio age and maturity change PV staffing and scaling

(09:50) - Biosimilar PV complexity vs. small molecules and novel therapeutics

(12:02) - PV leadership: building a team that functions as one unit, not a checklist

(13:52) - "We are education poor and training rich": training vs. education

(17:09) - Trust, explainability, traceability, and AI governance

(18:02) - What happens when it goes wrong: the UK air traffic control (NATS) outage

(23:35) - The aviation analogy: human in the loop, human in command(25:23) - From reactive to predictive safety: what "predictive" actually means

(33:19) - Technical vs. non-technical skills, and the CEO test

(35:43) - Pav's personal turning point: the PSUR, the CD-ROM, and "is this it?"

(43:15) - Pharmacovigilance 2.0, or already 3.0? The 5-year AI outlook

(51:23) - The biggest mindset shift: from insular thinker to systems thinker(55:37) - Closing remarks and thanks

Every pharmacovigilance system runs on a safety database — the central engine where every ICSR is entered, processed, and reported. But choosing, validating, and maintaining that database is one of the most consequential infrastructure decisions a biotech will make, and one of the most common sources of inspection findings when it is done badly.

In Episode 6 of What's Next in PV, Tereza Korecka speaks with Jan Kolouch and Vojtech Kvita from NextPV Services about how to approach the safety database decision end to end:

• What a safety database actually does inside the PV system, and when a company can function without one
• How vendor choice depends on ICSR volume, functionality, region, and budget — not just brand name
• Build vs. buy vs. outsource: control, cost, and who really owns your data in each model
• What GAMP 5 and EU Annex 11 validation actually requires: URS, IQ/OQ/PQ, and change control
• Practical setup steps: dictionaries, workflows, SOPs, and EudraVigilance gateway testing
• What triggers a database migration, and why it is one of the most error-prone PV projects
• Why due diligence, not the vendor's brand, determines whether the system holds up at inspection

This is a practical conversation for PV, QA, IT, and biotech leaders selecting, validating, or migrating a safety database.

Hosted by Tereza Korecka | Guests: Jan Kolouch and Vojtech Kvita, NextPV Services

(00:00) - Intro music and episode introduction(00:31) - Opening and guest greetings(00:41) - Today's topic: safety databases(00:58) - What a safety database does inside the PV system(02:16) - Do you actually need one? Volume, PSURs, and reporting obligations(03:26) - Overview question: which vendors do companies choose from(03:51) - Main selection factors: volume, functionality, region, and budget(07:45) - Large vs. small organizations: enhanced functionality vs. practicality(11:26) - Build, buy, or outsource: framing the decision(11:59) - The build option, database ownership, and vendor-managed models(16:46) - Why building your own database is rare in practice today(18:21) - GAMP validation: what the process actually looks like(19:00) - URS, FS, IQ/OQ/PQ, GAMP 5, EU Annex 11, and the EU AI Act(23:21) - Beyond validation: what companies trip up on at setup(23:52) - Database setup: dictionaries, workflows, SOPs, and EudraVigilance gateway testing(29:22) - Changing databases: what actually triggers a migration(29:46) - Migration scenarios: E2B exports, script migration, and data enrichment(34:27) - When migrations slip: deviations, mitigation, and inspections(36:54) - Closing advice for biotechs facing a database decision(37:34) - Operating model, timeline buffers, and post-launch monitoring(42:18) - Due diligence as the thread through the whole process(45:23) - Migration usually happens alongside the day job(46:20) - Why it helps to have an internal database point person(47:33) - Closing remarks

Keywords:
pharmacovigilance, safety database, Oracle Argus, ArisGlobal LSMV, Veeva Vault Safety, GAMP 5, EU Annex 11, database validation, database migration, EudraVigilance gateway, ICSR, PV system, biotech pharmacovigilance

Choosing a PV tool is no longer just a safety database decision. Computerized systems now sit across pharmacovigilance operations: safety databases, PSMF tools, QMS platforms, reporting layers, integrations, dashboards, automation, and AI-enabled capabilities.

In Episode 5 of What's Next in PV, Tereza Korecka speaks with Jan Kolouch and Vojtech Kvita from NextPV Services about how PV teams should approach computerized systems as strategic infrastructure, not just software.

The conversation covers:

• What makes a tool relevant for pharmacovigilance and GxP work
• Why companies should start with process, data flow, and intended use before vendor demos
• Strategic routes: current vendor modules, new vendors, integrations, APIs, and internal builds
• How scorecards, sandbox testing, and business requirements help decision makers
• Why cost should be assessed beyond licence price
• What validation means for configurable and cloud-based tools
• Where AI fits, and why automation and audit trails still matter
• How the approach differs for small biotechs and large pharma companies

This is a practical episode for PV, QA, IT, regulatory, and biotech leaders reviewing their technology stack or preparing for a regulated tool selection project.

Hosted by Tereza Korecka | Guests: Jan Kolouch and Vojtech Kvita, NextPV Services

Keywords:
pharmacovigilance, computerized systems, computerized tools, PV systems, GxP validation, safety database, PV technology, AI in pharmacovigilance, automation, audit trail, data integrity, PSMF, QMS, vendor selection, biotech PV

(00:53) - Opening and guest greetings(01:01) - Episode topic: computerized tools in pharmacovigilance(01:20) - Computerized systems are broader than safety databases(02:55) - How to decide whether a tool is used for PV activities(04:07) - Why PV system decisions are becoming more complex(04:29) - Digitalization, AI momentum, and cross-functional decision-making(07:17) - Main strategic options for better PV tools(07:41) - Current vendor modules, new vendors, integrations, APIs, and internal builds(11:04) - Why teams should invest time in tool selection and interoperability(12:30) - What to assess before looking at vendor options(13:16) - Demos, system inventory, data flow, and validation considerations(16:03) - Scorecards, business requirements, and system requirements(17:02) - How to compare functionality without ignoring cost(17:39) - Shortlisting vendors, sandbox testing, and pricing models(20:06) - Building a viable business case and assessing benefit-risk(21:38) - What validation means for GxP tools(22:18) - Configuration, implementation, cloud setup, and validation packages(26:04) - Why validation can range from simple testing to a full project(27:11) - AI trends without turning the episode into an AI validation deep dive(28:05) - Enterprise LLMs, local models, system-of-record data, and human oversight(33:06) - Automation before AI and why audit trails matter(35:32) - Practical decision process: from use case to governance(36:13) - Define use case, regulatory impact, data access, cost, validation, and ownership(39:54) - Small biotech vs large pharma: what changes and what stays the same(40:21) - Lean compliant stacks, fragmentation, vendor lock-in, and governance(43:53) - Final takeaway question: choosing PV tools in May 2026(44:37) - Computerized systems as strategic PV infrastructure(47:07) - Do not be afraid to move toward computerized systems and supervised AI(48:44) - Closing remarks