Analytical Development Leader · Durham / Raleigh, NC

Eric J.
Yearley,
PhD

Biologics Protein Therapeutics mAbs ADCs Antigens AAV Gene Therapy mRNA Vaccines

Thirteen years building the analytical control strategies that carry complex biologics, from vaccines and mAbs to mRNA LNP and AAV gene therapy, from bench through Phase 3 and into patients' hands.

Get in touch View experience
⊹ m7G AUG ⊹
Eric J. Yearley, PhD
Plate · 01 E.J.Y. / 2026
13+
Years of Experience
Leading analytical development for cutting edge biologic modalities.
10×
Tests per FTE
Productivity lift under my leadership. 14 tests/FTE at baseline, 220 at peak.
6
Production AI Tools
Agentic workflows I built myself, recapturing 300 to 910 senior scientist days per year.
01 / About

Building the analytical foundation for breakthrough therapeutics.

Bio

Throughout my career, I have built and led analytical development organizations responsible for characterizing some of the most complex therapeutic modalities in the pharmaceutical industry. From recombinant protein vaccines to cutting-edge mRNA-LNP platforms and AAV gene therapies, I've developed the analytical strategies that enable these life-changing treatments to reach patients.

My expertise spans the complete analytical lifecycle, from early method development through late-stage validation, technology transfer, and commercial implementation. I take pride in building high-performing teams, implementing digital transformation initiatives, and driving operational excellence while maintaining rigorous regulatory compliance.

With a PhD in Chemistry from the University of Toledo (2007) under A. Alan Pinkerton, specializing in X-ray Crystallography, and postdoctoral experience at Genentech, I bring a deep scientific foundation combined with practical industry experience across companies including CSL Seqirus, BridgeBio, Precision Biosciences, GlaxoSmithKline, and MacroGenics.

i.

Team Leadership

Led teams of 10+ scientists across multiple disciplines.

ii.

Regulatory Excellence

Authored CMC sections for IND, BLA, MAA, and IMPD submissions.

iii.

Digital Innovation

98% reduction in lab deviations through ELN implementation.

iv.

Vaccine Development

Contributed to early clinical development of Arexvy, GSK's RSV vaccine.

02 / Industry Pulse

AI in biotech & CMC.

Nature Biotech
Generative models for de novo antibody design move toward the clinic.
Five programs using AI-designed candidates now in Phase I, with binder discovery times reduced 4-8×.
↗ Read8 min
FDA / CDER
Draft guidance on AI-assisted CMC document preparation.
Comments open on expectations for model provenance, validation, and traceability in Module 3 submissions.
↗ GuidancePDF
BioProcess Intl
Automated peptide mapping pipelines cut release timelines in half.
Case study: paired LC-MS data with LLM-authored narrative for IND-stage characterization packages.
↗ Case study12 min

Updated weekly · curated selection · opinions are my own

03 / Expertise

Areas of expertise.

01

Mass Spectrometry

LC-MS, LC-MS/MS, peptide mapping, and intact mass for protein and oligonucleotide characterization. I also build my own MS tooling: spectrum decoding, XIC extraction, charge envelope finding, sequence mass matching, and deconvolution, applied to capping analysis and modification mapping for mRNA.

02

Analytical Control Strategy

ACS frameworks aligned with ICH Q14: CQA assessment, FMEA risk work, method lifecycle mapping across phase transitions, ATP drafting, and specification setting per ICH Q6B. Recently delivered a Phase 1 to Phase 2 transition deck using AI as a reasoning partner on FMEA and risk determination, not just as a drafting engine.

03

Scientist Builder & AI Tooling

Production portfolio of agentic workflows built with Claude Code and OpenAI Codex: analytical report generator, work instruction generator, MS utilities, raw chromatographic AUC integration, and template driven PowerPoint generation. Quantified recapture of 300 to 910 senior scientist days per year against 8 to 15 weeks of upfront build.

04

Throughput & Team Leadership

Took over an analytical testing group Sept 2023 and drove a 9× increase in monthly throughput and a 10× increase in tests per FTE while reducing headcount 13%. Peak month 1,800 tests at 220 tests/FTE. Process consistency tightened from 0.45 to 0.31 CoV. ~14,500 tests across 26 months.

05

Method Validation & Transfer

End to end method qualification and validation per ICH Q2(R2), with global CDMO transfers and multi site implementation. Includes bioassay work (cell based potency, ELISA, immunofluorescence, FACS) and molecular methods (qPCR, ddPCR) for mRNA and viral vector programs.

06

Digital Transformation

ELN and LIMS implementation, lab automation, and raw data integration pipelines that bypass vendor software dependence for non routine work. AUC calculation direct from chromatographic raw files reduced per trace time from up to 2 hours to ~1 minute, which moved integration out of the rate limiting column entirely.

04 / Experience

Professional journey.

Jul 2023 to Present

CSL Seqirus

Associate Director, Cell, Molecular, Separations & CMC Analytical
Holly Springs, NC
  • Direct 13+ member analytical development organization with 5 direct reports and full P&L responsibility
  • Co-delivered analytical control strategies for 3 clinical and 3 platform programs enabling critical product advancements
  • Achieved 10× increase in fragment analyzer assay throughput and 7× increase in potency assay throughput
  • Led ELN implementation reducing deviations from 150/year to 2 or 3/year (~98% reduction)
  • Justified >$1M annual CAPEX and headcount requests aligned with portfolio priorities
Sep 2019 to Jul 2023

BridgeBio

Associate Director, Principal Scientist
Raleigh, NC
  • Managed 11+ direct and indirect reports; served on CMC leadership team
  • Lead AD for novel pre-IND viral vector program, achieved IND-enabling tox material in 6 months
  • Authored CMC sections for multiple IND submissions and FDA annual updates
  • Co-led design of cGMP QC release testing facility for clinical batch release
  • Implemented QbD framework defining CQAs, CPPs, and comparability strategy
Mar 2019 to Aug 2019

Precision Biosciences

Senior Scientist
Durham, NC
  • Co-led development of assays to quantitate critical nucleic acid species for gene therapy programs
  • Developed assay to quantify 5′ capping species using reverse phase ion pairing LC-MS
  • Assisted in determination of CQAs and analytical techniques from QTPP
2016 to 2019

GlaxoSmithKline

Expert Scientist
Rockville, MD
  • Expert scientific and matrix leader for LC-MS, separations, biophysical and computational activities
  • Critical analytical contributor to early clinical development of Arexvy (RSVPreF3 OA), GSK's RSV vaccine
  • Co-led design of Good Laboratory Practice Program; authored 10 GLP guidance documents
  • Awarded two Silver Performance Awards for vital scientific contributions
2012 to 2016

MacroGenics, Inc.

Scientist I
Rockville, MD
  • Established and managed protein LC/MS program for bispecific antibody and ADC characterization
  • Led implementation of SEC-MALS and DSC technologies for enhanced characterization
  • Supported computational modeling for antibody design through analytical insights
05 / Impact

Case studies.

CSL Seqirus

Flow cytometry potency assay optimization.

A critical potency assay was creating a bottleneck in vaccine development, with 200-sample campaigns requiring 8 months to complete, threatening program timelines for late-stage advancement.

Through rigorous scientific analysis, my team demonstrated that 2 to 3 data points in the linear region provided statistically equivalent results to full dose-response curves. This insight enabled a complete workflow redesign, ensuring a critical initiative was finished within the established time frame while maintaining 98% Right-First-Time quality.

7×
Throughput Increase
8→2
Months / 200 Samples
98%
RFT Maintained
BridgeBio

Gene therapy analytical platform.

An AAV9 gene therapy program required transferring a comprehensive analytical platform to multiple CDMOs, including methods to confirm enzyme expression and demonstrate functional potency of the therapeutic protein.

Led multi-site technology transfer of the complete analytical suite: LC-MS methods for protein characterization, SEC and CE methods for purity, and a two-stage potency approach, first confirming transgene expression, then validating enzymatic activity. Achieved IND-enabling tox material in 6 months.

RMAT
FDA Designation
4
Special Designations
6 mo
To IND-Enabling
CSL Seqirus

Fragment analyzer platform implementation.

Late-stage testing demands required dramatically higher sample throughput than existing agarose gel electrophoresis could deliver: only 20 samples per week with low resolution that risked program timelines.

Led transition to high-resolution Fragment Analyzer technology, developing and validating methods that replaced manual gel-based workflows with automated capillary electrophoresis. The new platform enabled high-resolution nucleic acid characterization at scale.

25×
Throughput Increase
200
Samples / 2 Days
High-Res
Data Quality
06a / Throughput Ledger

Took over an analytical testing group and transformed its throughput.

Analytical Testing Group Leadership
Sept 2023 to Nov 2025 · 26 months · ~14,500 tests
Holly Springs, NC
Monthly tests
9×
Throughput Increase
Baseline ~131 tests/month rising to peak ~1,800 tests in Oct 2025.
Tests per FTE
10×
FTE Productivity
14 tests/FTE/mo at baseline, 220 at peak. 26 month average ≈ 58.
Avg headcount
−13%
Headcount Reduction
9.5 FTE down to 8.3 while throughput grew 9×. Average 9.6 FTE over 26 months.
1,800
Peak Month
October 2025 · 220 tests/FTE with 8.3 FTE.
~9,700
Final Year Volume
Up from ~1,570 tests in Year 1. A 6× annual lift.
1,400
Sustained · Last 3 Months
~180 tests/FTE/month. The new run rate, not a spike.
0.31
Coefficient of Variation
Down from 0.45 in the first six months. Process consistency tightened.
~14,500
Tests Run · 26 Months
~560/month average across the full tenure.

Figures from internal rollup, Sept 2023 to Nov 2025.

06b / AI Portfolio

A scientist builder portfolio.

Production · 2024 / 2025

AI assisted tools and agentic workflows that changed how I execute analytical development for mRNA LNP and epigenetic silencing therapeutics.

Over the past several months I built out a portfolio of production tools spanning document authoring, strategic deliverables, bench data interpretation, and software development more broadly. What follows is an inventory of what I built, how it's being used, and the quantified time savings where I have real usage data behind them.

I code the prototypes myself using Claude Code and OpenAI Codex in agentic flows. The MIT Sloan AI in Pharma & Biotech program keeps the frame grounded in real industry constraints.

01

Analytical Report Generator

Agentic authoring · Claude Code + Codex

Produces full reports across six assay types: mRNA purity, sgRNA purity, capping, poly A tail length, LNP lipid content ID, and RiboGreen. A first draft reaches roughly 90% completeness in about 15 minutes of generation. Finishing work takes another 2 days per report.

Baseline for an experienced scientist is 10 to 15 working days per report; for a junior author, 2 to 3 times that.

The coverage flip: reports that never got written for lack of runway now move from "would be nice if we had time" into "routinely produced."

Baseline
10 to 15 dSenior, per report
AI tooled
~2 dPer report
Reduction
80 to 95%
Recaptured / yr
150 to 450Days
02

Work Instruction Generator

Batched authoring · Shared template backbone

Produces work instructions for the same six analytical methods on a shared template backbone. Six WIs generate in parallel in 1 to 2 hours. Lab verification (running each procedure end to end) adds roughly 5 days; finishing and review, 2 to 3 more.

Baseline for six WIs authored individually is 30 to 54 working days for a senior author, double or more for a junior. Batching six simultaneously in 1 to 2 hours is qualitatively different from writing them one at a time, which was never practical because of template churn and context switching.

WIs now stay current with actual method practice instead of lagging by quarters.

Baseline
30 to 54 dSix WIs, senior
AI tooled
~8 dSix WIs, batched
Reduction
75 to 90%
Recaptured / yr
30 to 90Days
03

Analytical Control Strategy Deck

Phase 1 to Phase 2 transition · AI as reasoning partner

Built a full slide deck covering the analytical control strategy with AI used as a working partner throughout, not just as a formatter. I used Claude to construct an FMEA risk assessment for the phase transition, map the evolution of analytical methods across those phases, and draft ATPs for Phase 2. Claude guided me through the risk determination reasoning itself rather than simply rendering outputs. Timeline components were generated as Gantt charts via Claude Code, and I closed the loop by having Claude run a consistency review on the finished PowerPoint to catch misalignment across sections.

Baseline calendar duration is 8 to 10 weeks of heads down senior scientist effort. I finished in 4 weeks while spending roughly 60% of my time at the bench, which translates to about 1.6 weeks of effective focused effort on the deck.

This is the workflow that most clearly demonstrates AI as a reasoning partner on judgment heavy activities (FMEA, risk determination, method evolution) rather than a drafting engine on templated ones.

Baseline
8 to 10 wkHeads down, senior
AI tooled
~1.6 wkEffective effort
Reduction
80 to 85%
Recaptured / yr
30 to 80Days
04

Mass Spectrometry Tooling

Mass Parser · Capping Analysis Flow · Extended Capping

Rather than rely on vendor software for non routine MS analysis, I built a suite of in house utilities. Mass Parser bundles spectrum decoding, XIC extraction, charge envelope finding, sequence mass finding, and deconvolution. Capping Analysis Flow parses Agilent LC MS text exports, identifies and quantifies cap species against theoretical masses of the caps in use, and outputs Excel tables and PowerPoint plots automatically. The extended version adds UV integration, batch processing, multi replicate handling, and automated PowerPoint generation on top of the base flow.

The sequence plus modification permutation matching against experimental masses is particularly valuable for capping characterization and oligonucleotide modification mapping.

Exploring modification permutations by hand or in a spreadsheet is tedious enough that it often got abandoned. With the tooling, alternative interpretations take minutes, so the exploration actually happens.

Baseline
0.5 to 1.5 dPer analysis
AI tooled
< 1 hrPer analysis
Reduction
90 to 95%
Recaptured / yr
30 to 90Days, 30 to 60 analyses
05

Chromatographic AUC Integration

Raw data integration · Vendor software independent

I'm calculating area under the curve directly from raw chromatographic data using AI generated analysis routines, bypassing dependence on vendor software for non routine integration needs. This feeds the capping flows and sits alongside the assay report generators.

At 1 minute per chromatogram, integration is no longer rate limiting. Reprocessing historical data with updated parameters, parameter sensitivity analysis, and real time iteration on separation methods all become routine.

Baseline
15 m to 2 hrPer chromatogram
AI tooled
~1 minPer chromatogram
Reduction
95 to 99%
Recaptured / yr
12 to 50Days, 50 to 200 traces
06

Template Driven Presentation Generation

PowerPoint generator · Company templates + written instructions

A PowerPoint deck generator that ingests company templates and produces slides from written instructions, closing the loop between analysis output (Excel, plots) and the final communication artifact.

This is the workflow that compounds with every other workflow on the list. Every analytical output eventually lands in a slide for an internal review, steering committee update, or cross functional alignment meeting. Cutting the PowerPoint tax by 90% makes every downstream analysis faster in practice than its own per workflow number suggests.

Baseline
1.5 to 3 dPer 15 to 30 slide deck
AI tooled
1 to 2 hrPer deck
Reduction
90 to 95%
Recaptured / yr
50 to 150Days, 50 to 120 decks

Aggregate time recaptured.

Senior analytical scientist days, annualized
WorkflowDays / year
Analytical Report Generator150 to 450
Work Instruction Generator30 to 90
Analytical Control Strategy deck30 to 80
Mass spectrometry tooling30 to 90
Chromatographic AUC integration12 to 50
Template driven PowerPoint generation50 to 150
Benchling ELN integrationDeferred, in development
Sequencing pipeline (mRNA seq, methyl seq)Deferred, runtime pending
Total quantified~300 to ~910
FTE equivalent
1.5 to 2.5
Senior analytical scientist capacity recaptured per year, defensible midpoint.
Payback period
2 to 4 mo
Against 8 to 15 weeks of upfront tool building investment across the portfolio.
ROI · 2 to 3 yr horizon
10× to 20×
On direct labor. The recaptured capacity isn't showing up as shorter hours. It's showing up as bench time. Currently running about 60% lab time while executing senior analytical strategy deliverables that would otherwise pull me fully off the bench.
The coverage flip

How much faster · how much more science.

The time savings numbers above measure how much faster. The coverage number answers how much more science gets done, and the second one is probably the bigger win over the life of a program.

  • 01Analytical reports get written instead of deferred indefinitely.
  • 02Work instructions stay current with actual method practice instead of lagging.
  • 03Non routine chromatographic integrations get done properly instead of approximated in Excel or skipped.
  • 04MS characterization runs on every lot and timepoint instead of a sampled subset.
  • 05Modification permutation exploration happens routinely instead of being abandoned as too tedious.
  • 06Decks get polished instead of shipped rushed.
  • 07Analytical outputs communicate more frequently and in more digestible chunks, which drives faster program level decisions.
The meta point

No longer constrained by commercial software.

Any workflow, any functionality, any data analysis I need, I can build. I've become self sufficient as a scientist builder, creating custom software and analysis tooling tailored exactly to the analytical questions in front of me.

Most analytical leads at my level have the domain expertise but not the coding capacity to build tools like these. Most scientist coders have the coding capacity but not the analytical depth to know what tools are worth building. The combination of both, augmented by AI, is what makes this portfolio feasible. That's the positioning that matters for the next phase of analytical development work in cell and gene therapy, epigenetic silencing, and mRNA LNP therapeutics.

06 / Skills

Technical competencies.

Analytical technologies

RP-HPLCSECIEXHILIC LC-MSLC-MS/MSPeptide MappingIntact Mass qPCRddPCRELISAFACS CE-SDSSEC-MALSDSCDLSAUC

Product categories

Monoclonal AntibodiesBispecific AntibodiesADCs mRNA-LNPAAV Gene TherapyRecombinant Vaccines Viral Vectors

Regulatory & quality

cGMPICH Q14ICH Q2(R2) ICH Q6BICH Q5E INDBLAMAA ALCOA+GSP

Leadership & systems

Team Leadership (10+)CDMO Management ELN (IDBS)LIMS (Labware) AGILE / ScrumDigital Transformation

Independent AI development

PythonRAG Systems LLM IntegrationSynthetic Data Prompt Engineering

Education & certifications

PhD Chemistry · U. Toledo · 2007 MIT Sloan · AI in Pharma & Biotech · 2025
07 / Publications

Peer-reviewed work.

2023
Enrichment of Adeno-Associated Virus Serotype 5 Full Capsids by Anion Exchange Chromatography with Dual Salt Elution Gradients
Lavoie, R.A.; Yearley, E.J. et al., Biotechnology and Bioengineering
22 citations
2018
Non-Enzymatic and Site-Specific Glycan Shedding: A Novel Protein Degradation Pathway Observed in a Stabilized Form of RSV Prefusion F Protein
Qian, J.; Yearley, E.J. et al., Analytical Chemistry
18 citations
2014
Observation of Small Cluster Formation in Concentrated Monoclonal Antibody Solutions and Its Implications to Solution Viscosity
Yearley, E.J. et al., Biophysical Journal · 106(8), 1763 to 1770
215 citations
2013
Small-Angle Neutron Scattering Characterization of Monoclonal Antibody Conformations and Interactions at High Concentrations
Yearley, E.J. et al., Biophysical Journal
130 citations
2011
Shear-Induced Metastable States of End-Grafted Polystyrene
Sasa, L.A.; Yearley, E.J. et al., Physical Review E
8 citations
2010
The Couette Configuration of the Los Alamos Neutron Science Center Neutron Rheometer for the Investigation of Polymers in the Bulk via Small-Angle Neutron Scattering
Yearley, E.J. et al., Review of Scientific Instruments
16 citations
2010
The Los Alamos Neutron Science Center Neutron Rheometer in the Cone and Plate Geometry to Examine Tethered Polymers / Polymer Melt Interfaces via Neutron Reflectometry
Sasa, L.A.; Yearley, E.J. et al., Review of Scientific Instruments
8 citations
2008
Experimental Electron Density Studies of Non-Steroidal Synthetic Estrogens: Diethylstilbestrol and Dienestrol
Yearley, E.J. et al., Journal of Molecular Structure
44 citations
2007
Binding of Genistein to the Estrogen Receptor Based on an Experimental Electron Density Study
Yearley, E.J. et al., Journal of the American Chemical Society
87 citations
2007
Electron Density Studies of Selected Nonsteroidal Estrogens
Yearley, E.J., PhD Dissertation, University of Toledo
2007
5-Aminonaphthalene-1-sulfonic Acid and Its Manganese, Nickel and Cobalt Salts
Genther, D.J.; Yearley, E.J. et al., Acta Crystallographica
9 citations
2007
2,3,10,11-Tetrathiatricyclo[10.4.0.0⁴,⁹]hexadeca-4,6,8,12,14,16-hexaene
Yearley, E.J. et al., Acta Crystallographica
5 citations
Featured in
  • Chemical & Engineering News, "Why Monoclonal Antibody Drugs Get Sticky"
  • Department of Commerce, USA, "NIST Research Offers Guide in Formulating Cancer Treatment Drugs"
  • NIST, "NIST Helps Cancer Treatment Drugs Get Past Their Sticking Point"
  • Scientific Highlight, NIST Center for Neutron Research
  • Scientific Highlight, Institut Laue-Langevin, France
08 / Open Source

GitHub work.

github.com/eric-yearley · AI tooling, content, prototypes
podify
Public

AI powered content to podcast tool. Turns PDFs, URLs, Markdown, and plain text into natural multi voice audio with Claude generated scripts, OpenAI or ElevenLabs TTS, and RSS publishing built in.

PythonClaude APIOpenAI TTSStreamlit
greeting-card-app
Public

AI powered greeting card generator built on a bring your own API key model. Node and Express backend with Google GenAI for image and copy generation, deployable on Vercel.

Node.jsExpressGoogle GenAIVercel
eric-yearley.github.io
Public

Source for ericyearley.com. The site you're reading now. Editorial portfolio with animated biologics background, throughput ledger, AI tooling inventory, and the rest.

HTMLCSSGitHub Pages
View all on GitHub
09 / Contact

Let's connect.

Let's connect.

Always interested in discussing new opportunities in analytical development leadership. Biologics characterization, digital transformation, or building high-performing analytical teams.

LinkedIn ↗ GitHub ↗
Email Click to reveal email →
Résumé Download PDF ↓
Location Raleigh / Durham, NC

Tweaks×