LLM-Powered Research Discovery System
Overview
This project presents a full-scale system architecture designed to assist researchers in discovering relevant academic evidence based on natural language queries. It integrates advanced search indexing, large language models (LLMs), semantic similarity analysis, and a strong emphasis on ethical AI practices. The system takes a research question as input and outputs contextually relevant evidence extracted from a corpus of academic abstracts.
Description Objective: To design an intelligent system that enables researchers to retrieve and interpret academic evidence using AI, with a focus on accuracy, explainability, and ethical accountability. Core Technologies: PostgreSQL: Structured data storage for abstracts. Elasticsearch: Full-text and semantic indexing for fast retrieval. NLP & LLMs: Transformer-based models for deep contextual analysis Cosine Similarity: Semantic matching between user queries and stored abstracts. Python: NLP pipeline, query logic, and backend processing. Visual Analytics & FeedbackSystems: To ensure user engagement and system transparency. Key Features: Indexed and categorized academic abstracts using topic modeling and NLP. Implemented semantic similarity algorithms to match research questions with relevant literature. Designed a transformer-based language model layer to extract, summarize, and annotate key evidence. Built UX recommendations for an intuitive user interface and contextualized feedback loop. Integrated visual analytics and bias-monitoring to ensure ethical, inclusive outputs.
Fashion Venture
Smart City Innovation Framework
Overview
Red Deer 1M+: A Scalable Innovation Framework for Urban Transformation.
Creating sustainable infrastructure through Open Innovation 2.0, Living Labs, and SME empowerment.
This strategic innovation project presents a roadmap for transforming Red Deer, Alberta into a thriving, future-ready metropolis of one million people. The core challenge addressed is how to sustainably scale urban infrastructure while preserving livability and ecological integrity.
Key Components:
-
Open Innovation Hub: A collaborative platform where city stakeholders co-develop transport, housing, and green tech solutions
-
Living Labs: Real-time testing environments for electric mobility, smart grids, and inclusive design
-
Policy Design: Tax and procurement incentives for SMEs driving infrastructure and clean energy innovation
-
Impact Investing Models: Tools like community bonds and circular value capture systems to ensure all contributors benefit
This model is informed by global smart city best practices and adapted to Red Deer’s unique demographic, geographic, and governance context.
Open-Source Robotics Commercialization
Overview
Delcrobot: Building a Business Model for Open Innovation in Robotics
Transforming open-source into sustainable value through community, customization, and strategic adoption
This project addresses the challenge of commercializing a modular, open-source robotic system under the permissive MIT license. The goal was to design a business model that maintains the spirit of open innovation while enabling sustainable revenue generation.
Key Components:
-
Open Business Model Layer: Premium add-ons, consulting services, and targeted B2B applications
-
Adoption Layer: Documentation, support portals, and structured onboarding for hobbyists and enterprises
-
Crowdsourcing Layer: Community challenges and contribution loops to enhance system capabilities and market fit
This framework reflects how modern robotics ventures can thrive at the intersection of transparency, agility, and user-driven evolution.
