Datarize is a B2B SaaS company that provides data-driven CRM marketing solutions for e-commerce businesses.- Automation of personalized campaigns- Optimized onsite and messaging campaigns - Optimizing customer engagement based on customer behavior data.- Product Recommendations - User analytics from top to bottom of the funnel- Touch all parts of data from engineering to analysis to modeling Keywords: Data lake, Athena, MySQL, Pyspark, EMR, batch-data-pipeline, Data governance, Data lineage, DW setup and design

- Protopie is a design based SaaS company. My job as data scientist is to bring together growth, product and sales data together and solve business problems.- Data infrastructure setup: Data warehouse (Redshift) architecture setup from source to target using AWS ,Built data pipelines ETL services, logging of erroneous dataset.- Product, Marketing and Growth Data Science,Analytics- SaaS metrics: MRR/ARR/CPC/Lifetime value of customer/Churn Analysis/ Forecasting Problems/ Cohort Analysis - Customer segmentation algorithms, Lead Scoring, Churn prediction.- KPI and dashboarding- AWS data engineering (Lambda, Event bridge,appsflow, DMS, Glue, redshift, EC2, Crawler) - Corporate Strategy

A- IoT Platform(B2B):Creating IoT platform by automating the terminal operations and optimizing the operational performance (operational visibility and control on resource usage). As part of IoT, I am playing hybrid role of operational analyst and data scientist working mainly on timeseries, event and meta data.ETA (Estimated time of arrival): Developed a deep learning model (LSTM-CNN based architecture)-Deep TTE toestimate travel time for trucks at the port and recommend the optimized route. This helped port in saving 25% ofthe resources (fuel consumption, working hours, machine coordination, traffic congestion) used by findingbottlenecks in the operation• KPIs, dashboard and reporting of BI model (Microservices): Led from conceptual to completion stage of dashboardby defining key metrics of port operation. Key metrics included OEE, Terminal/Trucks productivity, containerwaiting time, reefer waiting time, container traffic, truck visit time, traffic congestion seasonality. This dashboardwas later implemented in further four ports. Data was obtained using IoT devices attached to machines andequipments.• Truck drivers dangerous driving behavioral modeling using data obtained from IoT devices attached to machines.• Equipment maintenance monitoring: I used anomaly detection model to find anomaly in speed, fuel and enginetemperature of the machines. This helped port in improving engine condition proactively. It was also used asperformance metric for truck drivers.• Database Schema design and ETL pipelines: Led a team to design, build, and refine existing PostgresDB schemaand timescaleDB. Created customized python scripting ETL pipelines to feed data into monitoring, operationaldashboard and Terminal Visualizer.

Worked mainly related to autonomous vehicles, driving behavioral analysis, traffic analysis, operational optimization domains, smart card based activity analysis. I played hybrid role of all three data phases (data engineering, data science/analytics) mainly working on geospatial, survey, timeseries, web scraped event dataset.1. Transportation Bus smart card data - Routing feasibility analysis, trip chain analysis, hot spot locations using R language.2. DTG data of taxi and bus - algorithm development to identify dangerous driving behavior location identification using R, map view library and statistical analysis. Visualization using Tableau.3. Autonomous bus - user before and after taking a ride behavior analysis4. smart traffic signal (coordinated traffic signals optimization) - Algorithm development using R and R shiny.5. Shortest path algorithm development incase of disaster/flooding: Implementation of DBSCAN algortihm and networkx,osmnx in python and dashboard development using tabpy in Tableau..6. Autonomous Bus speed data anomaly detection using LSTM-AE/Isolation Forest model using python and dashboard development using tabpy in Tableau.7. Improving public transportation network service by comparative analysis between trip chain behavior of public and network centrality using graph theory.

Researched and worked mainly related to user behavior analysis, transportation analysis, operational optimization and econometric analysis using Deep learning and statistical models.- Created a novel framework to analyze the decision-making behavior of travelers using Machine learning Copula based joint modeling. Knowing the inter-dependency and actual causal relationship would help policy makers and destination marketing organizations in coming up with unbiased policies. - Econometric Analysis/Machine learning.- Data driven approach for improving LOS of taxis (taxi demand prediction using CNN, bus travel time prediction using bus information system). - Prediction Problem -STTF (Short term traffic Forecasting using Deep Neural Networks (FCN, CNN, K-means clustering) and compared performance with traditional time series forecasting models(ARIMA/Average mean/Facebook Prophet).- TimeSeries Forecasting Problem- Merging behavior of Autonomous Vehicle at the merging road section (Drone analyzer to collect data and Reinforcement Learning to allow merging- Signal timing Optimization of highway in front of KAIST using CORSIM software.}

Monitoring and Evaluation Service:Tasked with monitoring and evaluation of fast track projects of CPEC (China - Pakistan Economic Corridor). I was given the task of handling the packages of E-35 (Burhan-Hasanabdal) and (Havelian-Thakot KKH phase II cpec). roadway covering a distance of 150 KM including culverts, bridges and tunnels.- Defined KPIs and created operational,financial and analytical dashboard related to monitoring and evaluation of highway projects. - Creation of simulated data using software and creating test dashboard to be deployed in the field later. - Safety guidelines and respective business decisions were taken based on the monitored report weekly, bi weekly. Off field visits were taken to cross check the progress reported and actual progress in the field. Monthly and Bi monthly meetings were done with stakeholders to get the understanding of client and contractors about KPIs shown on the dasboard.

-Inspected project sites of SERRA & PERRA (Abbotabad and Muzafrabad) to monitorProgress and adherence to design specifications, safety protocols and state sanitationstandards.-Estimated quantities and cost of materials, equipment and labor to determine projectfeasibility.-Visited project sites during construction to monitor progress and consult with contractorsCSB&CSW(Chinese contractors) and on-site engineers of NESPAK.- Analysis of surveyed dataset using excel and R to study the project feasibility.