Technological Innovator, Problem Solver and Solution Facilitator
Arkadiusz’s Portfolio
This portfolio is being remade, as I have developed considerably in the last two years. For now it might seem rather empty, but simply refer to my CV listed on LinkedIn. In this space I will start with showcasing my AI projects as part of my Artificial Intelligence Data Specialist Level 7 Higher Apprenticeship and the work I am doing with EQUINITI as it transforms and implements AI.
Project by Arkadiusz Kulpa
AUTOMATED EXCEL-BASED TEST SCHEDULE, WITH AZURE DEV OPS APIs AND A PYTHON SCRIPT FOR COMPUTE-INTENSIVE CALCULATIONS
Report by Arkadiusz Kulpa
APPLICABILITY OF AN AI CHAT ASSISTANT IN RESOLVING KNOWLEDGE COMPARTMENTALIZATION WITHIN A WORKPLACE
API Tests: here
This report aims to investigate whether AI Innovations can help manage complex operations of a company suffering from compartmentalized knowledge, where various teams need to rely on the broader business for information and guidance which is hard to come by individual specialists. An efficient and scalable cloud LLM application is proposed, where staff could query and receive responses that are grounded in facts derived from company’s data to which the GPT model has access.
Such innovations will necessarily be applied during the company’s ongoing transformation, part of which is to utilise AWS Cloud services and so the artifact developed in the second part of this report showcases a scalable AWS LightSail Instance providing a cost-effective VPS for hosting application’s WordPress front-end and PHP Back-end server. AWS API Gateways communicate with AWS Lambda serverless function which stores the OpenAI API Keys and communicates with the model, response fetched is then sent back to the front-end showing the entire end-to-end process required to host a secure LMM app over Cloud Technologies.
Data Security and Privacy is considered should OpenAI become a third-party data processor for the company and a Document store, LangChain, is outlined as a framework that could facilitate tokenization of company data and storage in a Vector Store, so that the query sent to OpenAI API contains relevant snippets of documents in the form of word embeddings and a simple demonstration of Chat GPT 4.0 Document reading capability is shown as an example of typical Use Case within the company.
This project demonstrates that LMMs work very well with cloud technologies and can be implemented with ease, providing access to cutting-edge technologies to any employee regardless of their technological acumen.
Full report: here
Project by Arkadiusz Kulpa
FRAUD DETECTION COMPLEX NEURAL NETWORK FOR FINANCIAL FRAUD DATA USING AWS SAGE MAKER CLOUD COMPUTING SOLUTIONS
This report considers how Machine Learning can provide a solution for the issue of fraud such that affects COMPANY as a Share registrar and Pension service provider.
In the first part of the report AWS Sage Maker is outlined as a potential cloud architecture for creating a fraud detection model, with the second part of the report showcasing a fraud detection deep learning artificial neural network artifact created using public simulated dataset of credit card transactions, which establishes the principles and approach needed to apply similar model for COMPANY specific problems and datasets.
Data and privacy remain one of the key considerations and a plan for safe implementation within COMPANY is presented, whereby the PROD data could be used within COMPANY’s own infrastructure and system to first generate an anonymized or even fully synthetic / simulated dataset. Once such a data model is acquired, it is uploaded to a Cloud Architecture of choice, where a model is trained, tested, validated, and adjusted until it produces satisfactory results. Such a model is then moved on-premises and can be used to tackle real-world scenarios using PROD data.
Implementation of such a fraud detection system would need to be divided into several stages, based on typical business use cases and complexity of the ML solution required.
First stage, which is showcased in the artifact using public dataset, would focus on transactions – as units for analysis – with the neural network looking at features of a transaction itself for patterns indicating fraud. Future stages / updates could see an Ensemble model (a model made up of many specialist models) take a holistic view of an account or entire dataset to spot Anomalies and/or patterns in the data and could be used to spot fraud type or fraudster gangs.
The artifact is developed in Google Collab using TensorFlow, with a baseline model contrasted against progressively mode advanced experiments, allowing for objective comparison of model’s improving ability to predict and generalize. Focus is placed on data pre-processing, which is shown to have strong influence on the model’s ability to learn patterns. The results achieved indicate that COMPANY will best be served by generating its own dataset and supplementing it with simulated data, where each feature can be thought through and controlled, so that the model has best chance to find patterns.
The volume of data is not at issue in such a scenario as Transfer Learning can be applied, where a small dataset can be used to fine-tune an established State of the Art model in the Fraud detection category to specialize it for COMPANY data.