Asif Akram

Engineer turned Data Scientist with 4+ years of experience currently working in E-Commerce Analytics at The ICONIC. Well-versed in using data analytics, data transformations/visualization, technical training, SQL/Python, Machine Learning algorithms, and Data Engineering pipelines to provide actionable insights, business strategies and customer experience improvements in the FMCG sectors.

I aim to create global business impact with my work related to product development using data as the forefront. Highly interested in roles pertaining to a combination of Data Analytics, Product Strategy, Machine Learning and Data Engineering.

Personal Accomplishments:

• The Resistance: Game of Deception – A key contributor to the codebase of a free online version of a popular card game The Resistance. Responsible for the addition of two new characters and gameplay features in the update Check out the game here! - https://www.proavalon.com/
• Sentiment Analysis on approximately 9 million tweets – Created a Convolutional Neural Network capable of classifying sentiment on tweets (positive/negative) with an 80% accuracy. Naturally, most tweets in Twitter are negative.
• AIAA UNSW Rocketry - Team lead in the UNSW Rocketry team tasked with buidling an autonomous model rocket for American Institute of Aeronautics and Astronautics USA. Responsible for the design of the payload delivery & autonomous landing systems.

Key Contributions at HelloFresh:

• Revamped the existing predictive algorithm for recipe scores using logistic regression to achieve 97% accuracy, leading to more precise recommendations for menus based on customer preference
• Pioneered a comprehensive analysis of the recipe databases of HelloFresh Australia and New Zealand, as well as EveryPlate to determine the utilisation of new and repeat recipes in menu catalogue, which prompted an initiative by Global to analyse all other markets
• Collaborated with other analysts to automate weekly, monthly, and quarterly dashboards to keep up to date with OKRs (Objectives and Key Results) using Databricks, building ETL pipelines using Spark instances, saving up to 6 hours per week.
• Developed an ETL/ELT pipeline for automation of recipe sequences for each customer preference based on live data of the type of recipes customers swap into, to be used as BAU for product planning team. This was a key part of an ongoing data-driven customer experience improvement process.
• Developed a comprehensive live dashboard in Tableau of the recipe databases of HelloFresh Australia, New Zealand, and EveryPlate outlining key metrics such as recipe scores, COGS, AIP, order uptakes, etc for stakeholders across APAC & Global
• Identified key contributors to customers pausing their subscriptions and developed a clustering prediction model (K-means) based on historical data to justify maximum constraints leading to significant cost savings upto $50k per market. This initiative sparked the attention of HelloFresh Global to scope possibility in their markets.
• Gathered data from local and global teams from the data warehouse using complex SQL and Python to create a single data source for internal teams to track KPIs and customer data for easy accessibility to build Post-Launch Reviews of concepts. Robust pipeline involves processing upwards of 50 million rows of data.

Key Contributions at Cundall:

• Provided a comprehensive HVAC map of all systems, ducts, and ventillation at Sydney Aiport Terminal 1 and Terminal 2, including a list of all Air Handling Units and their manufacturers. This allows for easier planning for expansion/changes to any future projects at Sydney Airport.
• Provided Mechanical ventillation and air conditioning plan for 23 public schools across NSW in 2 weeks, with a full heating/cooling load calculations and list of relevant suppliers of units
• Lead project engineer for the HVAC expansion of Marrickville Metro Shopping Centre, designing for aesthetics and sustainability.

Key Contributions at Abyss Solutions:

• Trained a comprehensive Machine Learning algorithm for detection of corrosion levels in assets on oil platforms. Clients included Total, Chevron & BP.
• Automated and ideated a streamlined process for writing inspection reports for corrosion, asset/fault detection and cracks on numerous underwater infrastructure.