Masoud Akbari Email & Phone Number

Tehran, Tehran Province, Iran

• Worked on developing an Enterprise Coding Assistant• Harnessed the power of different open-source and commercial LLMs• Deployed open-source LLMs such as Llama 2 and Mistral on GPU instances• Used RAG concepts to improve LLM responses and keep them in context• Conducted several prompt engineering which improved the accuracy of responses• Used Python and FastAPI for backend development• Participated in the design and implementation of the database• Utilized pgvector… Show more • Worked on developing an Enterprise Coding Assistant• Harnessed the power of different open-source and commercial LLMs• Deployed open-source LLMs such as Llama 2 and Mistral on GPU instances• Used RAG concepts to improve LLM responses and keep them in context• Conducted several prompt engineering which improved the accuracy of responses• Used Python and FastAPI for backend development• Participated in the design and implementation of the database• Utilized pgvector in PostgreSQL to include vector database• Participated in the deployment phase Using Docker and Kubernetes technologies Show less

Tehran, Tehran Province, Iran

• Used concepts of MLOps to develop customized ML pipelines• The developed pipelines facilitated the management of data and experiments• Utilized open-source tools to develop pipeline components• Developed pipelines for structured and unstructured data• Wrote clear documentation about different parts of the pipeline

• Worked with one of the provincial (Alborz province) gas companies to detect consumers with abnormal consumption behavior.• Set up a PostgreSQL database on our local servers to store client data when remote data access was impossible.• Applied different data exploratory analyses which helped gather insights from the data.• Used statistical and unsupervised approaches to make data cleaner, which improved anomaly detection results.• Implemented a deep self-supervised model to… Show more • Worked with one of the provincial (Alborz province) gas companies to detect consumers with abnormal consumption behavior.• Set up a PostgreSQL database on our local servers to store client data when remote data access was impossible.• Applied different data exploratory analyses which helped gather insights from the data.• Used statistical and unsupervised approaches to make data cleaner, which improved anomaly detection results.• Implemented a deep self-supervised model to detect abnormal consumption behaviors to deal with unlabeled data. Show less

Tehran, Iran

• Used Self-Supervised and Unsupervised concepts in a scientific paper for chatbots• Managed the creation of a Persian benchmark dataset for NLU tasks• Participated in the analysis of Topic Modeling algorithms on short Persian textsgathered from social media• Managed weekly meetings

Tehran Province, Iran

Master'S Degree, Artificial Intelligence, 3.8/4

Bachelor'S Degree, Computer Science, 3.03/4