Şevval Alper
Şevval es investigadora de IA en AIMultiple. Cuenta con experiencia previa en la generación de números pseudoaleatorios mediante sistemas caóticos.
Intereses de investigación
Şevval se centra en herramientas de codificación de IA, agentes de IA y tecnologías cuánticas.
Forma parte del equipo de evaluación comparativa de AIMultiple, donde realiza análisis y aporta información para ayudar a los lectores a comprender diversas tecnologías emergentes y sus aplicaciones.
Experiencia profesional
Contribuyó a la organización y orientación de los participantes en tres eventos de "Clases Magistrales Internacionales del CERN: física de partículas práctica" en Turquía, trabajando junto con el profesorado para facilitar el aprendizaje.
Educación
Şevval posee una licenciatura en Física por la Universidad Técnica de Oriente Medio.
Últimos artículos de Şevval
Prueba de referencia AGI: ¿Puede la IA generar valor económico?
AI will have its greatest impact when AI systems start to create economic value autonomously. We benchmarked whether frontier models can generate economic value. We prompted them to build a new digital application (e.g., website or mobile app) that can be monetized with a SaaS or advertising-based model.
8 modelos de código de IA evaluados: LMC-Eval
More than 37% of tasks performed on AI models are about computer programming and maths.
OCR Benchmark: Extracción de texto / Precisión de captura
OCR accuracy is critical for many document processing tasks, and SOTA multi-modal LLMs are now offering an alternative to OCR.
Comparativo de generadores de texto a video
A text-to-video generator is an AI system that turns written prompts into short videos by generating visuals, motion, and sometimes audio directly from natural language.
Ejecución de código con MCP: Un nuevo enfoque para la eficiencia de los agentes de IA
Anthropic introduced a method in which AI agents interact with Model Context Protocol (MCP) servers by writing executable code rather than making direct calls to tools. The agent treats tools as files on a computer, finds what it needs, and uses them directly with code, so intermediate data doesn’t have to pass through the model’s memory.
Top 10 Google Colab Alternativas
Google Colaboratory is a popular platform for data scientists and machine learning scientists, but its limitations and pricing may not meet your needs. Several alternatives offer unique features and capabilities that cater to different data science needs and scenarios.
LLM Parámetros: GPT-5 Alto, Medio, Bajo y Mínimo
New LLMs, such as OpenAI’s GPT-5 family, come in different versions (e.g., GPT-5, GPT-5-mini, and GPT-5-nano) and with various parameter settings, including high, medium, low, and minimal. Below, we explore the differences between these model versions by gathering their benchmark performance and the costs to run the benchmarks. Price vs.
Agentes de IA: Operador vs Uso del Navegador vs Proyecto Mariner
AI agents are increasingly marketed as end-to-end digital workers, but real-world performance can vary widely depending on the task, tools, and execution environment. To understand what these systems can genuinely deliver today, we conducted hands-on benchmarking across practical business scenarios.
Prueba de referencia de voz a texto: Deepgram vs. Whisper
We benchmarked the leading speech-to-text (STT) providers, focusing specifically on healthcare applications. Our benchmark used real-world examples to assess transcription accuracy in medical contexts, where precision is crucial. Speech-to-text benchmark results Based on both word error rate (WER) and character error rate (CER) results, GPT-4o-transcribe demonstrates the highest transcription accuracy among all evaluated speech-to-text systems.
Codificación por Vibra: Ideal para MVP Pero No Listo para Producción
Vibe coding is a new term that has entered our lives with AI coding tools like Cursor. It means coding by only prompting. We made several benchmarks to test the vibe coding tools, and with our experience, we decided to prepare this detailed guide.
Boletín informativo de AIMultiple
Reciba un correo electrónico gratuito a la semana con las últimas noticias tecnológicas B2B y análisis de expertos para impulsar su empresa.