European Journal of Science, Innovation and Technology

Architectural Patterns for Integrating Large Language Models (LLMs) into Node.js Server Applications

Oleksandr Tserkovnyi — Sun, 01 Feb 2026 00:00:00 +0200

This article examines the evolution and systematization of architectural patterns for integrating large language models (LLMs) into server applications built on the Node.js platform, against the backdrop of the rapid diffusion of generative technologies in industrial software development and the expanding market for Retrieval-Augmented Generation (RAG) solutions. The relevance stems from the fact that by 2025, LLMs will have become an indispensable component of digital products, while server architectures must embed computational speech into existing infrastructures under constraints of token budgets, call costs, and network latency. The objective is to identify and analytically describe stable architectural patterns that enable efficient, predictable LLM integration in Node.js backends. Methodologically, the work combines systemic architectural analysis, modeling of interactions with LLM APIs, and content analysis of industrial practices, enabling the author to construct an engineering-economic efficiency model for each configuration. The article’s novelty lies in formulating the concept of a balanced LLM-integration architecture in which throughput, token price, and service-layer observability are treated as interdependent architectural variables. An evolutionary pathway is proposed for transitioning from monolithic model calls to microservice and serverless patterns, informed by market growth dynamics and the scaling of compute resources. The article will benefit researchers and engineers engaged in server-application architectural design, cloud-service developers, and AI-engineering specialists aiming for resilient and cost-balanced deployment of LLM technologies in production environments.

Machine-Learning-Based Mapping and Ranking of Energy Materials in African Economies

Emmanuel Owoicho Abah, Christian Idogho — Sun, 01 Feb 2026 00:00:00 +0200

The global transition toward clean energy and advanced technologies has led to a rapid increase in demand for critical energy materials, including cobalt, lithium, rare earth elements, and platinum group metals. Although Africa possesses a significant share of these strategic minerals, the continent remains underrepresented in structured, data-driven mineral mapping initiatives. This research introduces a machine-learning framework based on artificial neural networks (ANNs) to predict and prioritize the likelihood of energy material occurrences across African nations. As demonstrated in a 2023 Nature Communications article, machine learning frameworks can map infrastructure such as distribution grids using publicly available multi-modal data, including street view images, road networks, and building maps. The results of this study confirm established mineral hubs, such as the Democratic Republic of Congo and South Africa, while also highlighting underexplored regions with substantial hidden potential. By addressing a critical data and strategy gap, this work provides a reproducible and scalable approach to resource intelligence, offering practical benefits for investors, policymakers, and researchers aiming to align African mineral development with the global energy transition.

A Proposed Master Development Plan for Conservation-Based Ecotourism at Tikub Lake, Tiaong, Quezon

Elzie T. Barbosa — Sun, 01 Feb 2026 00:00:00 +0200

Rapid tourism growth, expanding settlements, and resource-based livelihoods are exerting increasing pressure on many small freshwater landscapes in the Philippines. In sensitive crater-lake environments, even modest disturbances can trigger water-quality decline, shoreline erosion, habitat loss, and long-term ecological imbalance. These conditions underscore the need for planning approaches that safeguard natural systems while supporting local economic opportunities. This study responds to that challenge by examining the environmental, spatial, and governance conditions affecting Tikub Lake and formulating a conservation-based Master Development Plan (MDP) that aligns ecotourism development with ecological protection. The research evaluates the lake’s physical characteristics, land use patterns, stakeholder perspectives, and management issues to determine the requirements for a sustainable development framework. It also assesses existing zoning provisions, aquaculture practices, visitor behavior, and infrastructure gaps that influence the lake’s carrying capacity. The proposed MDP consolidates these findings into a structured set of planning strategies that include: (1) a spatial framework composed of a conservation core, protected shoreline strip, and ecotourism support zones; (2) refined land use and zoning designations to regulate development intensity; (3) access and circulation systems that minimize slope disturbance and manage visitor flows; and (4) facility planning standards for sanitation, waste management, trails, viewing areas, community markets, and low-impact eco-lodges. Together, these components create an integrated approach that strengthens environmental safeguards while enabling small-scale, community-driven ecotourism that is both economically viable and ecologically responsible.

A GIS and Space Syntax-Based Walkability Development Plan for The Heritage District of Taal, Batangas

Joseph Anthony A. Malabanan — Wed, 04 Feb 2026 00:00:00 +0200

This study developed a GIS- and Space Syntax–based Walkability Development Plan (WDP) for the Población Heritage District of Taal, Batangas, a heritage town where walking supports everyday mobility and tourism activity but is constrained by narrow streets, discontinuous sidewalks, and encroachments. Guided by local planning intentions that prioritize pedestrian connectivity while preserving heritage values, the research examined how spatial configuration relates to pedestrian movement and on-ground walkability conditions. A mixed-method convergent design integrated GIS-based spatial analysis, segment-based Space Syntax modeling, walkability audits, pedestrian counts, and perception surveys to assess pedestrian accessibility, safety, comfort, and infrastructure adequacy. Results show that pedestrian movement patterns align closely with spatial configuration: corridors with high global integration (Rn), high local integration (R3), and high choice (betweenness) correspond with observed pedestrian concentrations near major religious, commercial, and heritage nodes. However, audit and survey results indicate that many movement-critical streets exhibit poor walkability due to discontinuous or narrow sidewalks, obstructions, limited accessibility for persons with disabilities, and inadequate supporting amenities. The composite walkability index further indicates that very low walkability conditions occur in high-demand areas, clarifying the need for targeted and tiered interventions. Based on these findings, the study proposes a WDP that prioritizes corridors by urgency, classifies streets by functional typology, and outlines heritage-sensitive strategies for pedestrian space reallocation, traffic management, vendor regulation, inclusive accessibility, and monitoring mechanisms.

Bit-Width Quantization and Prompt Optimization: Achieving 90% Energy Savings in Large Language Models

Anupam Dhakal, Prashant Pokharel, Sabin Adhikari — Wed, 04 Feb 2026 00:00:00 +0200

For the rapidly evolving field of Large Language Models (LLMs, the rapid scaling has posed significant challenges. These problems include exorbitant energy consumption, prohibitively expensive deployment, and a significant impact on environmental sustainability. A major contributor to this problem is LLMs' colossal size. Typically, there are billions of parameters, and the need for them to be run in resource-scarce or edge environments. Our research delves into a functional and immediately applicable solution to kickstart the energy efficiency of LLMs by merging low-bit-width quantization and streamlined prompt techniques.

We have tested this approach with Llama-based models ranging from hundreds of millions to over one billion parameters and applied 4-bit post-training compression combined with structured prompt and query optimization to this spectrum of models. Utilizing a well-controlled A/B testing framework, we evaluated the task accuracy, delay, and power consumption between our baseline and optimized configurations. Since we can measure the actual power usage of our hardware, we could use the formula accuracy-per-watt to sum up the performance of both configurations. Our results show that 4-bit compression all by itself knocks out a significant portion of memory usage and electricity consumption, and then, our fine-tuning of the prompts cuts down the cost of token-level inference. When used in tandem, these two techniques have led to a 90% reduction in energy consumption with virtually no or statistically insignificant losses in accuracy on the tests we ran.

We also verified the effectiveness of this strategy for real-world use, demonstrating that it delivers consistent efficiency benefits when running on severely constrained hardware. The scalability analysis showed that this method still delivers a lot of bang for the buck even for models that have over a billion parameters.