About
Hi, I'm Andranik. I am a highly motivated problem-solver with a strong academic background, research experience, and award-winning projects. I have a demonstrated ability to understand multidisciplinary concepts and develop new skills quickly and efficiently. Being highly curious and passionate about learning, I welcome interdisciplinary projects as opportunities to expand my knowledge. My primary interest is working on agentic AI; I am particularly interested in building neuro-symbolic AI agents, leveraging the creativity and stochasticity of large language models while maintaining the robust, deterministic strengths of symbolic approaches. Feel free to connect with me on LinkedIn or reach out via email!
Resume
Professional Experience
Designed and implemented an AWS SageMaker ML pipeline for anomaly detection in financial accounting data, training personalized models for each customer and identifying anomalies using LOF + iForest + LLMs.
Full stack development building anomaly detection application, allowing accountants to create detection rules in natural language.
Implemented a POC for various generative AI use cases using open source LLMs (question answering over internal documents with RAG, Text-to-SQL for querying databases in natural language, summarizing patient app logs, etc.)
Designed and implemented a data lakehouse architecture POC using Databricks for large-scale, complex CRM device diagnostic data resulting in significant improvements to query performance, data storage, and access to low-level raw device data.
Designed and implemented tools/libraries for efficiently streaming, parsing, and redacting data from implantable cardiac device transmissions.
Developed tools and processes to decode raw device ECG data and create datasets for training machine learning models.
Developed ETLs, stored procedures, and designed schemas for OLAP databases
Data extraction, analysis, visualization, and reporting to support business-critical decisions, quality investigations, research/product development, and regulatory submissions.
Developed automated testing pipeline for mobile applications and medical devices.
Data analytics and reports for clinics and internal operations.
Internal tool development.
Skills & Technologies
Programming Languages
AI/ML Technologies
Cloud & Infrastructure
Frameworks & Libraries
Databases & Tools
Education
Focus - Data Science & Cloud Computing
Relevant Coursework - Advanced AI, Large-Scale Database Systems, Data Science, Cloud Computing, Quantum Computation
Focus - Intelligent Systems
Relevant Coursework - Machine Learning & Data Mining, Probabilistic Graphical Models, Reinforcement Learning, Graph Algorithms, Information Retrieval, Edge Computing
Research Experience
Algorithmic detection of structural variations to discover potential off-target activity during genome editing.
Analyzed Bos Taurus genome edited with TALENs.
Worked with FDA's MINI-HIVE HPC cluster for NGS analysis.
Documented and presented results to Dr. Simonyan's research team.
Research and development of novel compression algorithms for genetic data.
Developed visualization tool that creates a bitmap image representing the quality scores of a FASTQ file.
Implemented a novel Monte Carlo and Simulated Annealing-inspired algorithm that groups sequences with similar quality scores for compression.
Awards
Developed a mobile application using Flutter that incentivizes social distancing by using a geolocation tracking and gamification approach to award points for staying at home and away from crowded places, which can then be redeemed for discounts on services such as food/grocery delivery.
Built REST API and Admin Panel using Firebase, NodeJS, Express.
Top 20 finalist out of over 230 international teams composed of scientists, physicians, researchers, etc.
Developed a mobile application that wirelessly connects to a robotic physical therapy glove and assists patients with wrist mobility impairments in completing exercises.
Awarded $20,000 and 1st place out of over 1,000 national teams.
Pitched project at AT&T Developer Summit in Las Vegas.
Used C++ and Python for embedded device programming (Arduino/Bluetooth), Ionic for mobile web app.
