Aytekin Akbulut
Learner
(1)
4
Portals

Skills

Application programming interface (api) 2 Artificial intelligence 2 Database design 2 Innovation 2 Systems architecture 2

Achievements

Latest feedback

Recent projects

Work experience

Founder / Full-Stack Developer
Pythia
Toronto, Ontario, Canada
November 2024 - February 2025

● Pythia is a financial analysis app designed to help investment banks and individual investors assess a public company’s intrinsic value.
● Collaborated with a co-founder and seasoned financial advisor on marketing and business development to secure pre-revenue funding.
● Created a news aggregator that automatically gathers and analyzes company-related articles.
● Built the flagship feature, an equity research report generator that queries Pinecone indexes (annual/quarterly reports, tables, news), applies Piotroski-like metrics, and produces comprehensive analyses with references.
● Currently developing a neural network to extract features from these reports and predict future company performance.

Software Developer
PowerDev
Oakville, Ontario, Canada
July 2024 - November 2024

● Built and maintained high-performance APIs for efficient data retrieval and storage
● Ensured data integrity by validating and cross-referencing public datasets, implementing quality controls, and automating data checks.
● Leveraged .NET and Entity Framework to model large datasets and execute complex queries.
● Enhanced data accuracy by implementing quality control mechanisms that detect and correct data inconsistencies.
● Created extensions/utilities to standardize data attributes, reducing redundancy and improving modularity.
● Used Git-based version control to maintain code consistency and support agile collaboration.
● Integrated logging and monitoring to track performance, ensuring uptime and swift issue resolution.
● Automated data validation processes to ensure the integrity of imported news data, cross-verifying sources to maintain accuracy and reliability.
● Collaborated in an

Salesforce Developer
Soft Innovas
Toronto, Ontario, Canada
July 2022 - January 2023

● Configuring OWD, Permission Sets, Profiles, Role Hierarchies, and Field-Level Security for proper access.
● Creating custom objects, fields, formulas, record types, and relationships (Master-Detail, Lookup).
● Building validation rules, flows, custom layouts, approval processes, workflows, triggers, and sales processes.
● Developed efficient, budget-conscious solutions to meet client requirements.
● Translating client needs into Salesforce objects, lookups, junction objects, and master-detail relationships
● Implementing Apex Classes, Triggers, SOQL, SOSL, and thorough unit testing.
● Developing Lightning Web Components (LWCs) to improve efficiency and user experience.

Personal projects

AI-Driven Medical Diagnosis Assistant
January 2025 - February 2025

● Developed an AI-powered medical diagnosis application that predicted illnesses based on symptoms, test results, and medical images.
● Trained deep learning models on the ChestX-ray14 dataset to detect chest illnesses, leveraging CNN-based architectures for accurate medical imaging classification.
● Implemented natural language processing models to analyze clinical notes and patient symptoms, which provided diagnostic insights.
● Built an end-to-end pipeline for data preprocessing, model training, and deployment, ensuring efficient handling of large-scale medical datasets.
● Integrated vector databases to store and retrieve patient records, enhancing diagnostic accuracy through streamlined data access.

AI News Aggregator
June 2024 - November 2024

● Designed and developed a robust news API, that collects articles based on desired company, and industry.
● Integrated OpenAI API to classify and tag news articles, enabling precise topic categorization and improved content discovery for users.
● Implemented AI models for automated summarization, reducing article length while retaining key information for users who prefer concise content.
● Generated embeddings for the articles. Then stored the data in a Pinecone index to allow semantic search for future data analysis tasks.