Hello,
My name is Al Gharakhanian and have used Dataiku's platform extensively for rapid prototyping and other applications and have produced favorable results for my clients. Dataiku's platform has enabled me to reduce the development time significantly. Another less discussed advantage to Dataiku is its ability the unearth potential problems and obstacles at very early stages of a project.
In addition to Dataiku, I have extensive experience in developing, training, and deploying end-to-end Machine Learning models able to solve pressing business needs. I have experience in various use cases such as:
- Natural Language Processing (NLP)
- Time Series Forecasting
- Anomaly Detection
- Customer Churn Prediction
- Document and Text Processing
- Predictive Maintenance
- Reading Comprehension (based on SQuAD dataset)
I have significant experience utilizing Amazon Web Services (AWS) and have used resources such as:
- SageMaker
- Textract
- Lex
- Comprehend) extensively
- GPUs
I have used a wealth of most prominently:
- DNNs
- LSTM and BiLSTM
- Transformer-based Language Models (i.e. BERT, . . .)
- Linear and Logistic Regression
- Random Forest
- XGBoost
Commonly use the following development environment:
- Dataiku
- Tensorflow
- SciKit Learn
- numpy, pandas, python, Ax, Deequ
Pay Special Attention to:
- Data Visualization
- Data Preparation
- Security and Governance
- Bias
- Monitoring
I am thorough, detailed oriented, firm believer of program management (use Asana, Trello). My projects and blog can be found at www.cogneefy.com. I am based in Redwood City, CA. but have entertained many remote assignments. I am also open to assignments overseas needing onsite presence.
Al
al@cogneefy.com
