Pfizer Externship via Extern·🟢 Open to work
Completed the Pfizer Advanced externship building AI-powered document intelligence (OCR, RAG, Python), backed by a self-directed portfolio in SQL, Salesforce, and BRD/FRD requirements work.
Work spans the Pfizer Advanced externship (AI document insights, data extraction) and a self-directed portfolio in SQL, Salesforce, and BA requirements documentation.
.jpg)
Pfizer · ✅ Verified by Extern · ⏱️ In progress
Prototype AI-powered document intelligence with Pfizer—using OCR, LLMs, and RAG to automate real enterprise PDF workflows and build a standout portfolio project.
Completed the full SQLBolt curriculum and independently designed and built a comprehensive, cross-referenced SQL reference guide covering query fundamentals through advanced concepts, including JOINs, aggregate functions, execution order, and data manipulation (INSERT/UPDATE/DELETE). Authored origin
Completed the Pfizer Advanced externship building AI-powered document intelligence (OCR, RAG, Python), backed by a self-directed portfolio in SQL, Salesforce, and BRD/FRD requirements work.
I'm Monica Johnson, a Nashville-based operations professional moving into Business Analyst and Customer Success/Implementation roles after 6+ years at Amazon in high-volume, SLA-driven fulfillment operations. I'm currently in the Pfizer Advanced externship through Extern, building AI-powered document intelligence (OCR, RAG, Python) for real enterprise PDF workflows. Alongside that, I've built a self-directed portfolio spanning SQL, Salesforce, and BRD/FRD requirements work, backed by a We Can Code IT certificate and a Google Project Management certification.
Externships
Pfizer Advanced: AI-Powered Document Insights & Data Extraction Externship
In progress
Experience
Operations Associate
Amazon · 2019 – 2026
Software Development Apprentice
We Can Code IT · 2018
Office Administrator (Teller & Receptionist)
Ohio’s First Class Credit Union · 2017 – 2019 (Leave of absence May–Dec 2018 to complete software development trai
Education
We Can Code IT (WCCI)
Certificate of Software Development
Salesforce Trailhead
Admin Beginner Trail, Trailhead Explorer Rank · Class of 2026
Google Coursera
Project Management Certification
St. Martin de Porres High School
High School Diploma
Skills
✅ Verified by Extern · ⏱️ In progress
Prototype AI-powered document intelligence with Pfizer—using OCR, LLMs, and RAG to automate real enterprise PDF workflows and build a standout portfolio project.
The externship explored AI-powered document intelligence for enterprise PDF workflows using OCR, LLMs, and RAG. Work included researching foundational LLM concepts and translating them into practical explanations. Deliverables documented how these models function and how they integrate with document-processing pipelines.
I described how LLMs generate text by next-token prediction, how Transformer attention operates across input tokens, and how models move from large-scale pre-training to fine-tuning with human feedback.
I described how LLMs generated text by predicting the next token, how Transformers attended across input tokens, and how models were trained via pre-training on large corpora then fine-tuned with human feedback to shape outputs.
The project received a scanned page, applied median and Gaussian blur to denoise, used CLAHE for local contrast enhancement, and applied Otsu's thresholding to produce a clean binary image that improved text clarity for OCR.
I processed SDF PDFs with PyMuPDF to extract text and bounding boxes, applied heuristics to pull table content and key fields, and documented missed text and encoding issues in a deliverable report.
I ran all three OCR engines on the same scanned certificate, compared field-level accuracy, output structure, and confidence signals, and produced a side-by-side report. The analysis found PaddleOCR best at recovering address and paired labels/values, and recommended it with a confidence-based…
A multi-page pharmaceutical PDF was ingested and chunked by page, embeddings were produced with all-MiniLM-L6-v2, and a hybrid retrieval pipeline (BM25 + vector fused by Reciprocal Rank Fusion) with cross-encoder reranking was implemented. The submission documented model, chunking, retrieval…
Completed the full SQLBolt curriculum and independently designed and built a comprehensive, cross-referenced SQL reference guide covering query fundamentals through advanced concepts, including JOINs, aggregate functions, execution order, and data manipulation (INSERT/UPDATE/DELETE). Authored origin
Analyzed a synthetic Samsara customer dataset using SQL and Excel to identify high-probability expansion candidates among existing accounts, managed through a four-ticket Jira Kanban workflow. Published as a fully documented GitHub repository with complete analysis and findings.
Authored a complete BRD and FRD for a legal-SaaS conflict-of-interest enforcement system, including a full Salesforce-style CRM data model, business rules/trigger logic, and role-based permissions. Produced a supporting data dictionary and process-flow diagram; published as a documented GitHub repos
Authored a five-part training guide for HR administrators on international worker classification within an Employer-of-Record (EOR) platform scenario, including worked examples, common mistakes, and a condensed job aid. Published as a documented GitHub repository, completing a three-project portfoli
