Description
- Anomaly & Emergent Behavior Detection in real-time.
- Saved a lot of computing and human resource spent on anomalous data
- Used pretrained embedding like bert, glove and pre-trained models like spacy to train and detect PII fields.
- Trained a NLG (Natural Language Generation) model to generate data quality rules.
- Founded startup which summarises news in 60 words for arabic users.
- Grew company from idea to a user base of 10k.
- Developed a sophisticated language model to summarize arabic news at scale.
- Built a competent team of developers and digital marketer from scratch and trained them.
- Lead the product design, development team and Marketing Planning.
- Applied statistical modelling and data mining techniques to understand and predict app build deployment behaviour.
- Understand business needs and requirements to translate into conceptual designs
- Developed language model for topic extraction
- Maintained operations data from across the organization in hadoop data lake.
- Involved in developing and testing of the ETL strategy to populate the data in the dashboard from various source systems.
- Analyse covid-19 effect on ANZ’s operation and transactions for situation management.
- Developed MapReduce jobs in Python for log analysis, analytics, and data cleaning.
- Work with product managers to formulate the business problem.
- Built crop detection capabilities at scale using time series, multi-spectral and phenology features derived from remote sensing data
- Applied advanced machine learning models like CNN’s, RNN’s for crop detection, yield estimation and anomaly detection to improve the capabilities of detecting closely matched crops, inter-cropping, mixed-cropping and small farm sizes.
- Worked on creation of curated and cleaned crop signatures to detect crops at plot to state level while being region agnostic
- Fused multi-sensor satellite data for improving historic data analysis capabilities
- Crop stage prediction using time series forecasting and Recurrent Neural Networks
- Complete Machine learning development lifecycle using mlflow and aws sagemaker for model training, testing and deployment.
- Recommendation systems: Customization stage, order cart stage and personalization for customers. Increased business by 10-15%.
- Assume ownership and work continuously to improve them.
- Deep Learning for Sentiment Analysis, Intent Analysis, Entity Extraction, Text Classification, Semantic Similarity
- Derived a novel way to quantify, combine and productionized
- automated dish-tag creation
- dish-dish similarity
- customer-dish similarity
- Created and maintained gcloud backend and data pipeline.
- Build, test, deploy and validate recommendation engines and other ML models
- Designing and tracking new KPIs and implementation of smart algorithms.
- Skills:
- R
- Agile Methodologies
- Strategic planning
- ADF
- SLA
- Python
- Hive
- Sagemaker
- Machine Learning
- Elasticsearch
- Databricks
- Data Wrangling
- NoSQL
- Mlflow
- Data Visualization
- Tableau
- Pyspark
- Image Processing
- Flask
- Model Deployment
- Airflow
- SQL
