Machine Learning Data Scientist

Zanef Yusufali Nalakanth

Knoxville, TN - ***** 603-***-**** *****.*****@*****.***

Professional overview

Highly accomplished Senior Staff Data Scientist with over 15 years of experience in data science, analytics, and machine learning. Proven record of applying advanced analytical methods and causal inference techniques to drive product and business growth. I am a senior individual contributor with a strong track record of leading deep analytics and experimentation to uncover and scale high-impact growth opportunities. Skilled in SQL, Python, and big data platforms, with exceptional ability to translate complex data insights into strategic, actionable business decisions.

Technical Proficiencies:

Programming & Scripting: Python (Mastery), SQL (Mastery), R (Mastery), Scala Machine Learning & Deep Learning: Predictive analytics, Time Series Analysis, Regression Analysis, Statistical Modeling, Pytorch, Tensorflow, NLP, Neural Networks, Generative AI, Large Language Models (LLMs), Hypothesis Testing, Collaborative Filtering, LangChain, LangGraph, LangSmith, Causal Inference Big Data & Cloud Platforms: Apache Spark, PySpark, Hadoop, Hive, Kafka, Databricks, Azure Data Visualization & BI: Tableau, Alteryx, Business Intelligence,, Matplotlib, Seaborn Version Control: Git, GitHub, GitLab

Core Concepts: Data Engineering, Software Development, Debugging, Data Collection, Data Exploration, Data Cleaning, Data Modeling

Publications:

• LTE forecasting using the ARIMA model in comparison to the ETS model – USCC Tech Journal 2017

Conducted a comparative analysis of ARIMA and ETS models for LTE forecasting, definitively proving the superior accuracy and statistical significance of the ETS model's performance.

• Identifying clusters of mobile users for wireless network planning using Apache Spark – USCC Tech Journal 2019

Pioneered a methodology utilizing Apache Spark and machine learning-driven graph networks to pinpoint high-density mobile user clusters, directly contributing to more effective capacity improvements and improved customer satisfaction

• A framework for evolving and centralizing end to end ecosystem of a wireless network using binning method – USCC Tech Journal 2024

Designed a novel framework to assess the geographical impact of multilayered wireless network solutions, pinpointing and addressing areas of spectral inefficiency to optimize network performance and inform strategic network growth

EXPERIENCE

Senior Data Scientist (systems performance), US Cellular, Knoxville, TN Jul 2012 - Present

• Designed, implemented, and maintained end-to-end ML pipelines for spatial data digitization using Spark, Scala, and Python, reducing data error from greater than 4% to less than 0.1%, ensuring high data quality and integrity for downstream analytics. This demonstrates significant experience with data sourcing, cleansing, and preparation.

• Pioneered and deployed advanced AI-driven predictive models (N-HITS) for large-scale capacity forecasting using PyTorch, significantly reducing forecasting error and improving efficiency by 5% for a

$500M budget.

• Developed and deployed machine learning models to production environments: productionized a throughput prediction model using TensorFlow, deploying it via a REST API within a Docker container for scalable serving. This highlights experience integrating ML applications into production pipelines and working with deep learning frameworks.

• Conducted a comparative analysis of ARIMA and ETS models for LTE forecasting, definitively proving the superior accuracy and statistical significance of the ETS model's performance. This showcases expertise in experiment design and algorithm development.

• Pioneered a methodology utilizing Apache Spark and machine learning-driven graph networks to pinpoint high-density mobile user clusters, directly contributing to more effective capacity improvements and improved customer satisfaction.

• Analyzed customer survey responses using NLP to provide a customer experience score for the network, in order to improve customer experience. This demonstrates experience with unstructured data and libraries like spaCy.

• Served as a technical leader within the team, providing mentorship to junior data scientists and championing best practices for data-driven problem-solving by conducting training in SQL and R.

• Designed and architected a comprehensive Tableau-based exception reporting system, providing actionable insights for operational optimization and long-term strategic planning, improving data-driven decision-making. This highlights the ability to present data insights and recommendations to stakeholders.

Systems Performance Engineer, US Cellular, Manchester, NH Jul 2011 - Jul 2012

• Pioneered the implementation of in-house Netact reporting suite as admin as well as early adopter for network design and performance analytics

• Single handedly maintained and troubleshot Nokia GSM network in Maine included site maintenance, outage recovery and capacity forecasting.

• Conceptualized, designed and delivered fully automated report using excel to report all key performance statistics for 2G, 3G, 4G and 5G

• Created and maintained plan for mutli-band deployment across the network while managing band balancing and band offload

• Advised leadership on network issues and provided analysis during deployment of 4G network across New Hampshire and Maine.

• Assist Field Operations personnel with access to system performance data.

• Monitor and report status of major RF projects including cell build plan, carrier addition, plans and system capacity.

• Report and track RF incentive plan results.

• Develop reports of network configuration and improve system performance reports.

• Provide summaries and recommendations of new network software loads, including development of test plans, M4P, and result monitoring.

• Serve as the primary point of contact for RF Systems Performance Assistance Request with infrastructure vendors.

• Research and resolve specific network issues based on customer tickets to improve network performance

• Honored with multiple awards for performance in various roles as well as restricted stock benefits

• Supported and provided backup for systems performance data analytics alongside regular responsibilities Project Manager, Polytechnic institute of NYU, Brooklyn, NY Aug 2005 - Sep 2007

• Led a team of 25 to ensure thorough preparation and evaluation of applications

• Developed and implemented an automated method to cross-reference submitted applications, test scores, and documents, creating complete evaluation files

• Served as the main point of contact between academic department heads and student applicants

• Supported the dean in developing professional financial growth charts as part of regular responsibilities

• Organized and executed programs to assist new graduate students and streamline the registration process

• Communicated with potential applicants to encourage applications

• Created and maintained a comprehensive database of current and potential applicants

• Designed and managed programs to simplify the application process

• Streamlined the application process through redesign and process improvements

• Coordinated with various departments to ensure timely delivery of applications to program advisors Chief RF Design Engineer, PTC, Riyadh, KSA Aug 2005 - Sep 2007

• Advanced rapidly from Associate RF Engineer to Chief RF Design Engineer within two years

• Led a team of four engineers to develop cohesive network plans and establish regional oversight

• Earned certifications and performance bonuses for exemplary work

• Collaborated on the design and maintenance of an in-house automated database for performance management, capacity planning, and optimization

• Ensured spectral integrity to meet network requirements using an in-house database

• Developed and maintained spectrum plans, supporting regional engineers with optimization and troubleshooting

• Acted as the lead for in-building design and planning projects, as well as special initiatives

• Created, maintained, and delivered refined prediction models using Mentum Planet software EDUCATION

MS Financial Engineering - New York University Brooklyn, New York June 2011 MS Electrical Engineering - New York University Brooklyn, New York June 2009

Contact this candidate