30-second snapshot · open to work

I make support content AI-ready so customers self-serve, and support tickets go down.

  • Currently
    Content Specialist @ Meta
  • Experience
    15+ years in knowledge & content ops
  • AI
    AI-fluent since 2017 · LLM retrieval since 2025
  • Availability
    Open to work · Remote
Trusted by teams at
Meta
Microsoft
VMware
EA
Blizzard
Proven results
0%+
Support ticket deflection
Fewer tickets across VMware, Microsoft & EA
0→0%
Knowledge currency for AI retrieval
Highest currency score in org at Microsoft
0K
YouTube subscribers
Self-service support videos that funneled customers into the knowledge base
0%
Faster editorial reviews at EA, scaled at Microsoft
Via AI-assisted content QA at EA
Short on time? Jump to:See my impactCareer historyWhat people say
Open to work

Hi, I'm Marcela.

I specialize in AI-ready content that enables self-service and scalable knowledge operations that reduce ticket volume. I create content that customers love to use.

From nerdy communication with gamers on Blizzard community forums to managing Microsoft support content and running a 30k YouTube VMware tutorial channel, I make sure customers always get the right content, the way they want it and how they want it.

Connect on LinkedInView Career

My Expertise Clusters

AI Content OperationsSEO, AEO, governanceSelf-Service StrategyKnowledge Management, KCSProgram ManagementChange ManagementCXExcellence
Recent impact at
MetaMeta
MicrosoftMicrosoft
EA GamesEA Games
VMwareVMware
BlizzardBlizzard

How my work compounds into business impact

One initiative, a measurable chain reaction — from content I created to fewer support tickets and happier, self-sufficient customers.

I built the channel
0K
YouTube subscribers
Founded and owned a self-service video channel from the ground up.
It drove discovery
Traffic into the knowledge base
Videos funneled customers straight to self-service articles.
Customers self-served
Read content instead of raising tickets
People solved problems on their own before contacting support.
The business outcome
0%+
Ticket deflection
Measured impact across VMware, Microsoft & EA.
What kept it working at scale
0→0%
Knowledge Currency for AI Retrieval
Highest currency score in the org at Microsoft.

Goal: get our customer-support content surfaced inside AI answers — Copilot and other assistants. We had the authority and schema, but the content itself was outdated, so it wasn't being trusted or cited.

  • Audited the entire knowledge base to pinpoint stale, obsolete, and duplicate content.
  • Refreshed and rewrote articles so every answer was current and accurate.
  • Introduced AI tooling to make currency decisions data-informed and to de-duplicate content at scale.
  • Partnered closely with knowledge reviewers to sustain accuracy and currency over time.
  • Retired obsolete content so AI assistants only drew from trustworthy, up-to-date sources.
0%Faster editorial reviews at EA, scaled at Microsoft
The same AI-assisted QA approach made review cycles dramatically faster, how I sustained currency at scale (proven again at EA).
Working with AI right now, at Meta

AI is at the core of how I work

At Meta, I debug why AI chatbots retrieve the wrong knowledge, then I fix the structure, metadata, and governance underneath.

Not a recent pivot. I've been applying machine learning and LLMs to real content and support problems since 2017. Today I make knowledge machine-ready so AI systems can retrieve and trust it.

Since 2017AI-fluent, before it was mainstream
Since 2025LLM retrieval work at Meta
3 credentialsearned in 2026
What I'm doing now at Meta
AI Retrieval Optimization

Partner with Product & Engineering to improve how AI systems retrieve and assemble knowledge base content, tracing AI output failures back to gaps in structure, quality, coverage, or governance.

Machine-Ready Knowledge

Apply XML tag optimization, semantic chunking, and keyword enrichment to improve retrieval relevance, precision, and recall.

Core AI capabilities
Retrieval & RAG
Retrieval optimizationPrecision & recall tuningSemantic chunking
Machine-ready content
XML tag optimizationMetadata enrichmentLLM-ready architecture
AI visibility
AEOGEOAI answer optimization
Governance & quality
Knowledge governanceEvaluation frameworksAI-assisted QA
Track record: applying AI & ML since 2017
  1. Microsoft
    LLM-Ready Content ArchitectureMicrosoft

    Pioneered AEO, GEO, and AI visibility for customer-support content.

    69 → 94% content currency
  2. VMware
    GPT-Powered Feedback AnalysisVMware

    Used GPT to speed up qualitative feedback analysis of knowledge base articles.

    80% less analysis time
  3. Blizzard
    ML Toxicity DetectionBlizzard

    Annotated 10,000+ Overwatch chat logs to train content-moderation models.

    10,000+ logs annotated
  4. Blizzard
    Trained ML Translation (SDL Trados)BlizzardWhere it began · 2017

    Where it started, in 2017. Early English to Polish machine translation of gaming help articles failed on Blizzard's game-specific jargon. I annotated and corrected the translations and built a specialized gaming vocabulary that taught the engine how to translate on its own, turning unusable output into reliable machine translation.

    ~500 articles annotated · first ML work
I built this entire website with AI, every line of code, every layout decision. Because I don't just work with AI. I build with it.
AI credentials
Languages and Global Work
Managed content across 8+ languages at Microsoft5+ years remote
Polish (Native)
English (Fluent)
Cambridge
Cambridge Certificate
Business English (BEC)
Education & Certifications
University of Opole
Master's — Public Administration (EU Governance)
University of Opole, Poland - EQF 7
BPP University
Postgraduate Diploma — Digital Marketing
DMI & BPP University - EQF 7
Certifications
MTU
Designing Innovative Services
MTU
MTU
Digital Marketing
MTU
Harvard Business Online
Business Analytics
Harvard Business Online
CMIT
Leadership and Management
CMIT
Track Record

Career Timeline.

Partner with Product and Engineering to optimize how AI systems retrieve, assemble, and use knowledge base content — tracing AI output failures back to gaps in structure, quality, coverage, or governance. Apply XML tag optimization, semantic chunking, and keyword enrichment so content is modular, reusable, and structured for both human and machine consumption. Improve retrieval relevance, precision, and recall through validated content-level corrections, and contribute to evaluation frameworks that connect knowledge quality to AI performance and customer outcomes.

AI RAGRetrieval
XMLChunking
P/RPrecision & Recall
Social Proof

What Colleagues Say.

“She has deep knowledge of the content business... Her portfolio achieved the highest content freshness score of all Office portfolios.”

Senior Content Publishing Manager
Microsoft

“Marcela had built a reputation for being an extremely hard worker... she shows a positive approach and inspires others.”

People Manager
Blizzard Entertainment

“Extensive experience in community support, multimedia content, KCS, and content management. Highly recommended.”

Senior Support Product Manager
VMware / Bentley Systems

Testimonials originally published on LinkedIn

Community

DEI & Leadership.

EAEA·VMwareVMware

"I Am Remarkable" Facilitator

Delivered 15+ workshops (online and in-person) across EA and VMware. Consistently high attendance and outstanding feedback. Recognised by leadership for driving employee confidence and self-advocacy.

+500 employees empowered
BlizzardBlizzard

Women's Network – Communication Officer (Cork)

As Communication Officer for the Blizzard Cork Women's Network, I owned the SharePoint page and authored regular newsletters, keeping members engaged and informed. Strengthened community and amplified women's voices in gaming.

200+ members reached monthly
VMwareVMware

Lean-In Circle Founder (Cork)

Founded and led the first Lean-In Circle at VMware's Cork office. Created a safe peer-mentorship space for employees to share experiences, develop leadership skills, and support career growth. Sustained active membership over several quarters.

First at VMware Cork
Personal

Who I Am After Work.

Beyond work: award-winning macro photography · knowledge management blog · community singing.

Macro Photography

Award-winning work featured in the Irish Examiner. Capturing the unseen world up close.

View gallery→

Knowledge Management Blog

Thought leadership on knowledge management, AI, and content strategy.

Read blog→

Carrigtwohill Gospel Choir

Community singing that lifts spirits.