Aug 2023 – Sep 2024 · 1 yr 1 mo

• Technologically leading a team of 7+ Data Scientists and ML Engineers to drive the development of multiple AI solutions to support smart-product requirements of QBO Advanced and Payroll - a series of products tailored to serve the accounting needs of Mid-Market businesses. Some of these initiatives include real-time forecasting of financial statements, automated entity mapping for invoice customization, typographical error detection, financial insights through optimized search algorithms and causal modeling, etc.

Oct 2020 – Aug 2023 · 2 yrs 10 mos

• Developed and deployed several AI/ML solutions:
• Autonomous Insights: a tree-based and gradient-based search approach for accurate, interesting, and confident insights from any structured data,
• Data Co-pilot Insight Assistant: a GenAI solution to enable high-precision QnA from any domain-specific data,
• Real-time propensity prediction through extensive clickstream modeling,
• Hyper-personalized peer grouping for financial benchmarking insights, etc.

Jul 2020 – Oct 2020 · 3 mos

Jul 2019 – Jul 2020 · 1 yr

• Tech lead in a team of data scientists and engineers to build state of the art in-vehicular collision detection technology using smart phone sensor data.
• Collaborating with the stakeholders to build Model Building pipeline.
• Building a temporal window based feature engineering framework to extract features from noisy smartphone sensor data.
• Creating ensemble of Boosted tree based and Deep Learning models to classify events.
• Collaborating with the product team and engineers for deployment.
• Mentoring other data scientists and data science interns.
• Working as a Consultant across other data science teams/projects in Zendrive, e.g. Driver vs Passenger trip classification using mobile sensor data, Sensor fusion based driver fingerprinting, etc.

Aug 2017 – Dec 2017 · 4 mos · Raleigh-Durham, North Carolina Area

• Working with Dr. Qiana R. Cryer-Coupet as a student consultant for a project on social science. The goal of this project is to determine the effect of parental behavior on sons' involvement in future delinquent activities in fragile families.

Jan 2017 – Jun 2019 · 2 yrs 5 mos · Raleigh-Durham, North Carolina Area

• Machine Learning Researcher in an interdisciplinary team at Laboratory for Analytic Sciences (https://ncsu-las.org):
• Analysing FSI (http://fundforpeace.org/fsi/), the present global standard for state fragility, and developing an alternative solution that eliminates its biases; relies on better data; and utilises more sophisticated statistical techniques.
• Automated compilation of data from various open-sources, e.g. WorldBank, GDELT, ACLED, UNHCR, COW, INSCR etc.
• Building a ML-based solution (penalised GLM, SARIMA etc.) to forecast state instability by creating a dynamic probability map for fragile states.
• In my PhD dissertation:
• Developed a state of the art algorithm for spatial prediction of weather elements using low-quality, noisy but dense spatial data coming from sensors installed in mobile devices as well as high-quality but sparse weather station data.
• Introduced Veracity Score for large spatial data, proposed a Veracity Score based robust methodology for inference in geostatistics.
• This work, entitled as "A Statistical Analysis of Noisy Crowdsourced Weather Data", has been published in Annals of Applied Statistics. It has won the best student poster award at SRCOS SRC 2017, student paper competition travel award at IISA 2017, registration award at SDSS 2018 and travel scholarship award at SRC 2019.
• 2. Theoretically validated the merits of the Veracity Score based methodology using spatial asymptotics and simulations.
• 3. Developing a scalable copula-based covariance model to flexibly estimate anisotropic covariance.

Aug 2015 – Dec 2016 · 1 yr 4 mos · Raleigh-Durham, North Carolina Area

• My responsibilities include:
• 1. Holding weekly office hours for doubt clearing sessions for engineering and business students.
• 2. Grading the answers for homework assignments and finals.
• 3. Posting the solutions, codes for lab sessions and weekly homeworks.
• 4. Proctoring and other responsibilities.

Jun 2014 – Jul 2014 · 1 mo · New Delhi Area, India

• Working under supervision of Prof. Deepayan Sarkar on Stochastic Modelling of Human Reference Genome Sequence. Using the sequence model to find the binding motifs in a focused sampling way.

May 2013 – Jul 2013 · 2 mos · United States

• Working under Padmashri Prof. Jayanata K. Ghosh to compare model selection strategies under misspecified model setup.