Alber Makaryan

Operate and extend distributed AWS workflows orchestrating end-to-end case processing; diagnose failures and inconsistencies across service boundaries. Automate data collection and statistical reporting on AWS Redshift and Power BI; analyze production logs in Splunk to trace failures and surface anomalies.

Deliver a semester curriculum on relational databases, SQL, and PostgreSQL, with labs in data modeling, indexing, and query optimization. Built the data collection and processing pipeline behind a computer-vision study of student engagement, running pretrained models on an HPC GPU cluster over more than 2 million frames.

Analyzed large support datasets to surface anomalies, investigating irregularities in the underlying data before drawing conclusions. Built automated reporting pipelines and dashboards.

Evaluated large language model outputs against structured correctness criteria and curated prompt–response datasets for supervised fine-tuning, involving systematic close inspection of data quality at scale.

Built automated collection systems in Python across many heterogeneous sources via web scraping, each with its own format and failure modes. Developed ETL pipelines on Apache Airflow normalizing inconsistent source data into a single unified schema, and investigated data gaps, format drift, and silent collection failures to root cause and harden the pipeline.

Pursued a Master of Science in Data Science, including an exchange semester at San Jose State University in 2024.

Earned a Bachelor of Science in Economics from Yerevan State University.