Pradeep Pasupuleti Email & Phone Number

Chicago, Il, Us

Santa Clara, California, Us

1. Sewer Pipeline Defect Detection Product: Saved 32% time spent in image detection, classification and localization of defects in the inner surface of sewer lines from videos captured by robotic rover camera.2. EV SaaS Platform: Increased utilization by 10%, minimizing downtime by 4%, enabling ease of EV charging using queuing and charge optimization. 3. MLOps SaaS Platform: Automated cloud agnostic development, deployment, management and governance of MLaaS models as APIs with varying combinations of ML backend and frontends. Reduced MLOps TAT by 15%.4. Predictive Driver Behavior: Reduced operational costs by 10% and improved driver safety by 40% by deriving actionable and contextual driver behavior risk score for fleet and insurance companies.5. Mining Operations Optimization Platform: Increased $8.5M net ROI over 5 years and increased utilization by 1.8% by modelling load haul cycle operational data from mining assets, derived insights that improves overall productivity.6. Heavy Equipment Predictive Maintenance Product: Reduced downtime by 6% by using survival models, lifetime data and run to failure data to predict remaining useful lifetime for machine parts, simulation of optimal replacement warranties. 7. Mine logistics Optimization Simulation platform: Trajectory-based Discrete Event Simulation and Process Mining platform improved tonnage productivity by 3% and helped efficiently investigate a wide range of proposed mine design logistic scenarios in mine planning, management, and optimization. 8. Retail Demand forecasting Product: Improved ROI by 17% by analyzing product consumption, causal and seasonal factors to forecast retail demand accurately and optimize supply chain and promotions costs.9. Clinical Trials Simulation and Prediction Product: Identify and recommend key factors influencing enrollment, predict site suitability risk for conducting clinical trials across geographies, cohorts and medical conditions. Reduced enrollment costs by 5%.

Founded Datatma a consulting firm with a mission to simplify Big Data/Advanced analytics and unravel business value. Successfully developed products, implemented projects, scaled the organization.Hospital Chain in Florida: Developed a Real-time Code Blue Prediction Platform ingesting streaming data from biomedical devices through a series of outlier detection and predictive models such as log regression, decision trees and random forests, to alert clinicians to important changes in patient conditions—so they can stay one-step ahead of code blue conditions like sepsis and heart attack.Middle east Government Social Insurance company: Headed the Advanced Analytics program at a government insurance agency. Maximized the response against Fraud across key customer interaction points. Created the advanced Analytics center of Excellence. Top Bank in US: Hadoop Data Lake Engineering, led the architecture and implementation of a petabyte scale Big Data Lake at a leading bank in US. Integrated the existing data DWH/BI infrastructure with the new cluster and implemented re-usable code using PIG and HIVE to perform data managementIntelligent Information Lifecycle Management: Chief Architect for an ILM data appliance using Hadoop, TIKA and Elastic Search as core components. This product is being used at one of the largest bank in USA. It ingests structured and unstructured data into HDFS and performs a deep discovery about the content, value, lineage, provenance and usage pattern of the data. This discovered metadata is derived by TIKA/Elastic search and custom handwritten algorithms such as Latent Semantic Analysis. The metadata is used to configure and fire automated business rules to perform next best action on data such as, Send data to upstream systems, Archive data or defensibly dispose data.

Us

Pioneered and headed the Big Data Analytics practice where I architected and developed enterprise scale Big Data analytics solutions. Built and scaled Big Data Analytics teams from scratch. Focused on using Big Data and Analytics to align with organizational strategic roadmap and demonstrable tactical gains. • Built a Data Lake using Apache stack. Created data pipelines using Flume, Sqoop, WebHDFS, HIVE and PIG to ingest, integrate, profile and clean multi-structured data from more than 40 data sources including mainframes.• Created the enterprise strategy by envisioning the Big Data business cases, implementing COE design and delivering measurable proof of value.• Benchmarked cluster metrics that integrate existing data DWH-BI stacks with Apache Stack. Configured various data types and volumetric.• Architected operational processes for data profiling, data quality, workflow, scheduling, audit. Established Data Governance procedures to safeguard sensitive data.• Extensively used ensemble techniques like Random forests and Ada Boost for modeling prediction of readmissions using claims data and hospitalization data.• Applied Decision trees (CART, c4.0) and Bayesian inference to claims data to predict fraudulent claims and incorporated the scores into new business rules.• Used Kmeans clustering algorithms for finding patterns and created detailed segmentation to better understand the impact of various socio-economic trends on Claims over time. • Used Bag of Words, NLTK, NLP algorithms to parse Claims and Social data. Implemented topic modeling with Latent Dirichlet Allocation to build a cognitive model that understands the textual relationships.• Integrated Data Lake with HP Autonomy to ingest huge volumes of claims data and build a segmentation index. • Used similarity ranking in Map Reduce algorithms for data de-duplication.

Hyderabad, Ts, In

Zigbee Real-time Data Analytics: Incubated, scaled and led the Advanced SW division for Atlas Copco, a heavy engineering giant, where I designed and developed a C++ based real-time analytics platform on Linux hardened PDAs. This platform ingests and analyzes real-time data to support split second decisions for controlling Zigbee enabled Robotic Arms. To achieve this, extensively used Active Queue Scheduling algorithms and Minimum Spanning tree based decision trees implemented in multi-threaded producer-consumer queuing containers.Wind Data Analysis Framework: Built the Site Suitability Automation framework using R and C++ for Clipper. This framework crunches raw data of met towers from all across the country. Used complex Optimization algorithms to calculate Wind Shear, Turbulence Intensity, Excess Fatigue calculation, Tip Deflection and Wake effects etc. This tools objective is to act as a decision making system for the site analyst to know whether a geographical and wind profile is within defined bounds for wind energy production.

Pune, Maharashtra, In

Successfully built and led end to end customer solutions where I played the roles ranging from pre-sales to delivery manager. Responsible for overall account growth. Delivered multiple concurrent projects. Participated in formulating future business development plans and strategies, Proposed innovative engagement models. Mobile Number Portability solution: Led and delivered a multimillion Mobile Number Portability solution for TRA, UAE, where the solution was implemented on a Microsoft stack (SQL, BizTalk and C# engine). The core logic of number portability was implemented using Onward routing algorithms. These algorithms are optimized to meet the strict SLA adherence required from the telecom regulator.Successfully delivered full lifecycle telecom system integration development projects for Microsoft/Broadsoft/TRA UAE/ ETISALAT/DU. Improved the processes around knowledge management, competency building, solution offering development to build key talent and intellectual leadership for my group.

Pune, Maharashtra, In

Bangalore, Karnataka, In

Led the team from to architect and build key components of Spinnaker application Made great inroads for accounts growth in the Shipping/Logistics vertical. Was successful in building the client account and driving customer satisfaction. Shipping Optimization Solution: Led the team that developed core C++algorithms for the product Spinnaker at Tideworks-Seattle. Implemented the Greedy optimization algorithms which provides shipping terminal operators the ability to increase cargo volume and reduce vessel turn time, while maximizing efficiency. Implemented the knapsack and travelling salesman algorithms to automate container location assignments to maximize space utilization and avoid costly re-handles. Eventually these algorithms help in automatically sequence discharge containers and calculate the most efficient load-back sequence to reduce vessel turn times.

In

GIS routing: Developed the RouteviewPro product whose core routing engine used C++ based Dijkstras algorithm and STL data structures to find shortest path between points of interest. Ported and optimized this engine to Windows CE and created one of the first in-car navigation products.

Post Graduate Program In Data Analytics And Optimization Cpee, Machine Learning And Big Data- The Program Is Ranked 3Rd In The World By Cio Magazine.

Bachelor Of Technology (Btech), Computer Engineering