Utilized data science and machine learning techniques to analyze pharmaceutical literature, uncovering correlations between ingredients and indications, and generating & summarizing reports for pharmaceutical companies and the TGA.

Skills:

• Python
• ElasticSearch
• Pytorch
• Huggingface
• LLM
• LangChain
• CiCD

I Executed Seq2Seq (command-to-query) conversion projects, information retrieval and recommender systems at this company.

Technologies:

• NLP(Seq2Seq, GPT2, T5, RoBERTa, ...) accuracy more than 95%
• Python(Tensorflow, Transformer, HuggingFace, Pytorch, Data gathering ...)
• MLOps(Docker, Aws, Azure, MLflow)
• SQL and MongoDB

Senior Software Engineer and Co-Founder

We was a startup in insure-tech domain. We developed some of products.

• Developed fraud detection system and rules extraction system in insurance claims with an accuracy higher than 80% with imbalanced data.
• Forecasted the risk of car insurance sales.
• Developed smart API management software for integrating different systems.
• Led a team of developers and designers to build a scalable platform.
• Collaborated with stakeholders to define product requirements and prioritize features.
• Managed project timelines and budgets.
• Conducted code reviews and mentored junior developers.

In this company, with the trust of the manager, I was able to work efficiently as a developer and data scientist.

In web development field I developed health insurance software and API Manager application.

Technologies:

• C#
• SQL
• Redis
• ASP.NET

In Data Science I worked on customer loyalty, risk detection, fraud detection, sales forecasting, customer clustering and name entity recognition.

Technologies:

• Python
• TensorFlow
• Database warehouse

Member of Natural Language Processing Laboratory at University of Tehran.

Summary:

• Development a Aspect-Based Sentiment analysis with Deep Learning. (BERT, Word2Vec, ...)
• Persian data collect (Digikala and Snappfood customers opinions)