Sample Report — Illustrative  |  For Demonstration Purposes Only
Perlustrate
AI Readiness Advisory
Engagement type: Full Stakeholder Assessment
Delivery: May 2026
Confidential
AI Readiness Assessment Report
AI Readiness Assessment
Helios Biosciences
May 2026  ·  Prepared by Perlustrate

Assessment Overview

8
Stakeholder Interviews
2.0 / 5.0
Overall Maturity Score
13 hrs
Avg. Weekly Time Tax per Stakeholder

Helios Biosciences presents a profile consistent with a mid-stage biotech in Phase 2 clinical development: strong scientific talent and leadership enthusiasm for AI, constrained by fragmented data infrastructure and the absence of enterprise-wide governance. Across 8 interviews spanning R&D, Clinical Operations, Regulatory Affairs, Finance, IT, and Commercial, the dominant theme is manual effort at every data boundary — CRO reconciliation, cross-functional report assembly, and undocumented handoffs.

The overall maturity score of 2.0/5.0 places the organization in the Developing tier. Every dimension scored below 2.5, indicating a consistent pre-integration data environment. The primary readiness gap is not talent or motivation — it is infrastructure. The relative strength of Organization & Culture (2.3) signals that internal conditions for change are present, and that structured investment will find a receptive organization.

Assessment Approach

This assessment was conducted using Perlustrate’s structured interview protocol, delivered via AI-facilitated stakeholder conversations. Each session followed a standardized question set covering data workflows, tooling, governance awareness, AI exposure, and strategic priorities. Responses were scored using a five-dimension maturity framework aligned to AI readiness prerequisites.

Maturity scores are assigned on a 1–5 scale per interview, per dimension, based on evidence in the transcript. Dimension scores represent the mean across all 8 completed sessions (40 individual scores). Pain points are extracted and deduplicated by theme; frequency reflects the number of distinct interviewees who surfaced each issue.

Stakeholder Interviews Conducted

Name Role Function Duration
Dr. Margaret TranChief Scientific OfficerC-Suite/VP12 min
James OkonkwoDirector of Clinical OperationsOperations11 min
Sandra KowalskiDirector of Regulatory AffairsRegulatory/QA10 min
Dr. Kevin ParkSenior Research ScientistLab/R&D9 min
Priya NairClinical Data ManagerData/Analytics10 min
Rafael MendezIT Infrastructure LeadIT/Engineering8 min
Linda HorvathFinance ControllerOther9 min
Thomas ReyesCommercial & Business Development DirectorCommercial/BD8 min

Dimension Scores

2.0  / 5.0
Overall AI Readiness: Developing
Across all five dimensions · 8 interviews · 40 scored observations
Target threshold for AI deployment readiness: 3.0 / 5.0
Data Infrastructure
1.8Foundation
Point solutions only; Benchling isolated; CRO data requires manual intervention every transfer
AI/ML Readiness
1.9Foundation
Shadow ChatGPT use at leadership level; understands prerequisites; blocked by infrastructure not intent
Analytics & BI
1.9Foundation
GraphPad and R used by some; no BI platform; board reporting manually assembled from 5 sources
Data Governance
2.2Developing
No formal ownership or dictionary; GxP awareness without GxP-grade data governance
Organization & Culture
2.3Developing
CSO is a strong champion; scientists open; culture is willing but infrastructure isn't ready

Recurring Pain Points

Issues below represent themes surfaced independently by multiple interviewees. Frequency indicates the number of participants who raised each issue without prompting. Verbatim quotes are drawn directly from interview transcripts.

Data InfrastructureHigh severity
Data siloed across departments with no central source of truth
“the moment data needs to cross a functional boundary it basically falls off a cliff”
Raised by 7 of 8 interviewees
Analytics & BIHigh severity
Manual data assembly for reports and presentations
“I'd estimate I spend about thirty hours every quarter on that process. Which is absurd.”
Raised by 7 of 8 interviewees
Data GovernanceMedium severity
Lack of formal data governance and ownership policies
“that's kind of by osmosis rather than by policy”
Raised by 5 of 8 interviewees
AI/ML ReadinessMedium severity
Shadow AI use with no governance or data protection guardrails
“tolerated, but there's no formal policy”
Raised by 5 of 8 interviewees
Data GovernanceMedium severity
SharePoint as de facto document management with poor search
“SharePoint is a mess. Nobody really knows what's authoritative”
Raised by 4 of 8 interviewees
Data InfrastructureMedium severity
No unified view of clinical trial timelines and milestones
“finance will show one enrollment number and the CRO portal shows a different one”
Raised by 3 of 8 interviewees
Data InfrastructureHigh severity
CRO data reconciliation and report inconsistencies
“every data transfer arrives as a ZIP of CSVs with inconsistent column names”
Raised by 3 of 8 interviewees

Technology in Use Across the Organization

Tools cited across all 8 interviews. The count badge indicates how many interviewees mentioned each tool. Faded chips represent single-person mentions.

SharePoint11
Word8
Outlook8
Teams8
Excel8
PowerPoint5
Veeva Vault5
Benchling4
ChatGPT4
PubMed3
R3
Medidata Rave3
Zoom2
GraphPad Prism2
SciFinder2
Dotmatics2
DocuSign2
SAS2
Pinnacle 212
NetSuite2
Adaptive Insights2
Evaluate Pharma2
AWS S3
Adobe Acrobat
Azure AD
CDISC standards
CTMS (IQVIA)
Claude
Cortellis
Crowdstrike
EMA portals
Exchange Online
FDA.gov
GitHub Copilot
GlobalData
Jupyter notebooks
LinkedIn
Matplotlib
Medidata Rave EDC
Okta
Perplexity
Power BI
Python
Salesforce
Salesforce CRM
Seaborn
ServiceNow
Veeam

Research Grounding

Findings from this engagement are contextualized against current industry benchmarks and regulatory guidance relevant to life sciences AI readiness.

Regulatory Landscape
FDA’s guidance on AI/ML-based Software as a Medical Device (SaMD) continues to evolve, with the predetermined change control protocol (PCCP) framework now a central expectation for AI-enabled products in clinical workflows. Findings from this engagement — particularly the gap between R&D AI experimentation and production-grade deployment — align with the regulatory compliance gaps most commonly cited in FDA audit observations for mid-size biotechs.
Source: FDA Draft Guidance on AI/ML SaMD, updated Q4 2025.
Data Infrastructure Benchmarking
The Pistoia Alliance’s 2024 Data Maturity Survey found that 67% of biotechs with 100–500 employees operate at maturity level 2–3 on a 5-point scale for data infrastructure — consistent with the findings in this engagement. The most common gaps: absence of a unified data warehouse, siloed laboratory information systems, and lack of automated data quality monitoring.
Source: Pistoia Alliance Data Maturity Survey, 2024.
AI Adoption Patterns in Life Sciences
Life sciences organizations are adopting AI at approximately 1.3× the rate of other regulated industries (McKinsey State of AI, 2024), driven by competitive pressure in drug discovery and clinical trial optimization. However, 71% cite data readiness as the primary barrier — not model availability. This reinforces the Phase 1 prioritization in this engagement’s roadmap: data infrastructure investment precedes AI capability deployment.
Source: McKinsey Global Institute State of AI Report, 2024.
Recency Note
Research context reflects sources available as of Q1 2026. AI/ML regulatory guidance in life sciences evolves rapidly; the practical recency half-life for SaMD guidance is approximately 18 months. Findings should be re-contextualized at the next assessment cycle (recommended: annual).

Recommended Phased Investment

Recommendations are sequenced by dependency. AI capability deployment (Phase 3) requires data infrastructure and governance thresholds to be met in earlier phases. Each phase includes the minimum preconditions for advancing to the next.

1
Months 1–6
Foundation: Data Infrastructure & Governance
  • Audit and catalog all data sources: Benchling, Medidata Rave, Veeva Vault, NetSuite, SharePoint
  • Implement a cloud data warehouse as the single source of truth for cross-functional reporting
  • Establish automated CRO data ingestion to eliminate manual ZIP-file reconciliation
  • Deploy a formal data governance framework: data ownership, data dictionary, and classification policy
  • Publish an AI usage policy covering approved tools, data handling, and 21 CFR Part 11 boundaries
2
Months 6–12
Build: Analytics Capability & Governed AI Pilots
  • Deploy enterprise BI platform connected to the data warehouse; automate board reporting
  • Build self-service dashboards for Clinical Operations, Finance, and R&D milestone tracking
  • Pilot governed AI use cases: regulatory document drafting assistant and clinical enrollment monitoring
  • Instrument data quality monitoring and SLA alerting across CRO pipelines
  • Conduct cross-functional data literacy training (R&D, Clinical Ops, Finance)
3
Months 12–18
Scale: Enterprise AI Deployment
  • Deploy AI-enabled document intelligence for IND and submission assembly workflows
  • Implement ML-based clinical trial site performance and enrollment prediction models
  • Build self-service scientific data access: ELN queryability and experiment reproducibility tooling
  • Expand AI governance to cover model validation, bias monitoring, and audit trails
  • Conduct annual maturity re-assessment to benchmark progress and recalibrate roadmap