Pfizer Externship via Extern
I turn complex workflows into reliable platforms, backend systems, business-insight dashboards and AI systems.
A concise collection of class projects and practical work that show coding, problem solving, and readiness to contribute in a real-world environment.
Built a full-stack agentic AI platform using Python, FastAPI, Next.js, PostgreSQL, and Docker, transforming natural-language shopping goals into structured, budget-aware product recommendations. Designed a multi-agent LLM orchestration architecture for planning, product discovery, semantic evaluatio
Architected a full-stack agentic AI healthcare platform using Python, FastAPI, Next.js, LangGraph, and PostgreSQL for appointment booking, symptom triage, medical-report interpretation, reminders, and queue tracking. Built stateful multi-agent workflows with LangGraph, structured LLM routing, interr
I turn complex workflows into reliable platforms, backend systems, business-insight dashboards and AI systems.
Externships
Pfizer Advanced: AI-Powered Document Insights & Data Extraction Externship
In progress
Experience
Lead Software Engineer
Logiic Software Services Ltd | Kigali, Rwanda · Nov 2022 - Jan 2025
Education
University of British Columbia | Okanagan, Canada
Bachelor's degree in Computer Science · Class of 2027
Institut de Formation Apostolique de Kimihurura | Kigali, Rwanda
Diploma in Computer Science · Class of 2023
Microsoft
Azure AI-900 Professional Certificate · Class of 2025
IBM
Data Analyst Professional Certificate · Class of 2025
Skills
Built a full-stack agentic AI platform using Python, FastAPI, Next.js, PostgreSQL, and Docker, transforming natural-language shopping goals into structured, budget-aware product recommendations. Designed a multi-agent LLM orchestration architecture for planning, product discovery, semantic evaluatio
Architected a full-stack agentic AI healthcare platform using Python, FastAPI, Next.js, LangGraph, and PostgreSQL for appointment booking, symptom triage, medical-report interpretation, reminders, and queue tracking. Built stateful multi-agent workflows with LangGraph, structured LLM routing, interr
Architected and led full-stack development of Rwanda’s national school inspection and quality assurance platform, deployed across hundreds of schools for NESA. Engineered offline-first data collection and synchronization workflows, enabling reliable inspection processing in low-connectivity environm
Built a full-stack web application for digital school accreditation workflows across Rwanda, reducing application processing time by 50%. Implemented email verification, document management, real-time status notifications, and digital certificate issuance with 99% notification delivery success.
✅ 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 prototyped an AI document-intelligence pipeline that combined OCR, LLMs, and retrieval-augmented generation to automate enterprise PDF workflows. Work included researching model behaviour, designing integration patterns, and producing artifacts that demonstrated document parsing and semantic retrieval.
.jpg)
I produced a technical account of how LLMs tokenize input, map tokens to numerical vectors, model token relationships with transformer layers, and improve via pretraining and iterative fine-tuning, applied to pharmaceutical document processing.
I described how LLMs tokenized input, converted tokens to numerical vectors, used transformer layers to model token relationships, and generated responses. I noted pretraining on large corpora and iterative fine-tuning with feedback to improve accuracy.
