Rohit Agarwal
About
Detail
Senior Data Scientist at Lyric | Demand Forecasting | Supply chain | MLops
Bengaluru, Karnataka, India
Contact Rohit regarding:
work
Full-time jobs
Résumé
Jobs
verified_user
0% verified
Senior Data ScientistLyricOct 2025 - Current (11 months)- Senior Data ScientistantuitaiSep 2023 - Nov 2025 (2 years 3 months)* Spearheaded the delivery of advanced demand forecasting, pricing optimization, & supply chain solutions for three key clients, driving substantial improvements in operational efficiency and accuracy. * Implemented cutting-edge demand forecasting models using Time Series ESM, LightGBM, & LSTM, resulting in a 10% increase in forecast accuracy and more reliable inventory management across client portfolios. * Implemented dynamic store level cluster feature in forecasting models which in turn improved the forecast accuracy by 5%. * Designed robust evaluation frameworks to assess model accuracy, reliability, & robustness, and iteratively optimised forecasting models based on real-time feedback and business insights. * Designed robust evaluatio
- Data Scientist ||antuitaiJun 2021 - Aug 2023 (2 years 3 months)* Developed & deployed scalable demand forecasting solutions. Proficient in leveraging time series analysis, LightGBM, and sequence-to-sequence Neural networks to optimize forecasting accuracy and efficiency. * I was recognized for achieving 30% faster pipeline execution times and improving average demand forecasting accuracy by 8%. Moreover, I was appreciated by managers & peers for delivering results in a short duration by building different demand forecasting pipelines from scratch.
- Data Science AnalystMobile Premier League MPLMar 2021 - Jun 2021 (4 months)* Improved strategic analysis accuracy by {20%} through developing KPI dashboards for real-time gaming performance monitoring. * Developed a model using the KNN clustering algorithm to identify potential inactive users, improving customer retention by 4%.
- Data Scientist - CoELandmark GroupJul 2019 - Feb 2021 (1 year 8 months)* Led the Center of Excellence team, spearheading the delivery of innovative data science solutions to transform the apparel & furniture business operations. * Developed & implemented a workforce recommendation engine to optimize staffing across retail stores, enhancing operational efficiency. * Designed & deployed a markdown optimization engine, enabling dynamic pricing strategies at the SKU level to maximize sell-through rates. * Promoted the adoption of innovative data science practices within the organization, fostering a culture of continuous improvement and excellence. * Received a flash award at Landmark Group for developing impactful solutions.
- Data AnalystPharmeasyJul 2018 - Jun 2019 (1 year)* One of the first data analysts to develop data pipelines and data-oriented solutions for addressing the critical logistics challenges. * Played a key role in transforming the logistics vertical through data science, driving operational improvements and strategic advancements. * Developed and implemented models to optimize delivery boy payouts, ensuring fair compensation while aligning with business objectives and operational efficiency. * Created real-time alerts for anomalies and potential fraud in delivery operations, enhancing the overall customer experience and contributing to cost savings. * Ideated and developed various data models for ease of doing RCA & establishing a single source of truth.
- Data AnalystCozevaOct 2017 - Jun 2018 (9 months)* Developed predictive models to identify high-risk patients, enabling proactive interventions and personalized care plans to improve patient health and reduce costs. * Utilized SQL, Spotfire, and other tools to manage and analyze large datasets, enhancing the reliability and accuracy of data-driven insights. * Ideated and implemented data models to streamline (RCA) processes, establishing a single source of truth for operational data. * Applied (NLP) techniques to extract valuable information from unstructured medical documents, improving data accessibility and insight generation.
- DInternshipDefence Research and Development OrganisationJun 2015 - Jul 2015 (2 months)Web developer - Developed an admin site for keeping track of data of all the users of DRDO, which includes features like Password validation(for both users & admin), Deletion, Addition, Report view format date-wise, etc.
Education
verified_user
0% verified
- PGDM, Business AnalyticsPraxis Business SchoolJan 2016 - Dec 2017 (2 years)
- TBachelor's Degree, Computer ScienceTechno India JobsJan 2012 - Dec 2016 (5 years)
- TBachelor's Degree, Computer ScienceTechno IndiaJan 2012 - Dec 2016 (5 years)
- HHigh School, Computer ScienceHem shella Model SchoolJan 2010 - Dec 2012 (3 years)
Projects (professional or personal)
verified_user
0% verified
- HHEALTH CARE ANALYTICS (CLINICAL DATA)Jul 2017 - Sep 2017 (3 months)A clinic is visited by many patients. They usually have diabetes, hypertension, stroke and other heart diseases. The clinic wants to extract patterns from these patients’ data to devise a system which will tell the probability of having these diseases.
- EEffective cross selling of productsApr 2017 - May 2017 (2 months)Target- To find out which products are purchased together as a group of items, which can be used to strategize, cross selling activities. Approach- Used Association rule Mining to see what is purchased with what.The item matrix derived from the service variable was fed to the appriori algorithm with minimum 10% support & 50% confidence.The Association rules were then sorted with lift values & analyzed accordingly for bulding of services, increasing basket size of customers.
- TTo understand the cause of faliure of the Ariel liquid detergent- Research proposalApr 2017 - May 2017 (2 months)Objectives- * Conducting a brand health research, to estimate Awareness, Trial, etc in key markets and sub-groups of relevant target group, which should clearly diagnose the problems that the brand is facing. * To find percentage of customers migrating from Ariel detergent to Ariel liquid. * To understand the perception of the product among the Ariel liquid detergent users. The link of the proposal is attached above.
- PPredicting credibility of participants asking for loan using logistic regressionApr 2017Target - Analyzed customer's loan pattern & validated the eligibility for loan. Approach - Exploratory research was follwed by imputation of missing data.Thereafter, creating the logistic model with 5-fold cross validation.Henceforth the test data was fed to the model & the result of the data was analyzed through confusion matrix by evaluating certain parameters like accuracy, kappa value & ROC.
Awards
verified_user
0% verified
- DFlash AwardData Labs Pvt LtdOct 2019Awarded for developing the staffing algorithm & delivering phase 1 of the solution within a month's time. One of the kind developed from scratch with almost no usable data yet overcoming that restriction & provided a scientific solution that will impact the Centrepoint business both in terms of incremental revenue & cost savings. Along with the mathematical aspects the model also takes care of business requirements & realities.
- JJMOJMONov 2009
Publications
verified_user
0% verified
- KSanskrit as a language for next gen computersKolkataApr 2014Sanskrit is the first language to have a formal grammar defined; and there is evidence that Pāṇini’s work in this area influenced modern linguists like de Saussure and Chomsky. (And oh, Devanagari is awesome) One guy in NASA in the 80s tried to push Sanskrit as an ideal language for Artificial Intelligence applications; he was neither able to convince the AI community of this, nor was he able to make much headway in this himself. This approach is largely dead, but Indian media and the ancient-Indians-were-the-best crowd did not get the memo.