Jun 2023 – Aug 2023 · 2 mos · Los Angeles, California, United States · Hybrid

• 1. I spearheaded a scientific research project focused on improving invoice processing efficiency by leveraging Form Recognizers. Through meticulous experimentation and analysis, I successfully implemented real-time account code suggestions, resulting in a remarkable 71% reduction in processing time compared to manual coding. This achievement, facilitated by my expertise in product mapping and artificial intelligence, revolutionized our invoice processing workflows, enhancing overall productivity and accuracy. Additionally, I worked on a GUI based on Streamlit for real-time analysis of natural language processing data, enabling deeper insights into vendor and client behavior and further contributing to the 71% processing time savings.
• 2. Additionally, I played a pivotal role in reducing prediction error margins by 0.6x through the strategic application of natural language processing (NLP) strategies and hierarchical classification programming. Leveraging my proficiency in data science methodologies, I implemented innovative approaches to enhance model observability and accuracy, ultimately optimizing our predictive analytics capabilities.
• 3. Furthermore, I led the implementation of a streamlined ERP invoice processing system, which significantly improved accuracy and efficiency in account code suggestion while eliminating the need for manual Chart of Accounts (CoA) lookup. Through meticulous documentation and collaboration with cross-functional teams, I ensured seamless integration of this system, enhancing data processing pipelines and driving operational excellence within the organization. Additionally, my efforts in optimizing the invoice data pipeline, utilizing technologies such as Oracle databases and AWS S3 infrastructure, resulted in a substantial 71% reduction in processing time, further streamlining operations, and improving overall efficiency.

Feb 2023 – Jun 2024 · 1 yr 4 mos · Boston, Massachusetts, United States · On-site

• 1. As a Research Assistant at Northeastern University, I developed and implemented a cutting-edge machine learning solution for analyzing underwater mammal sounds across various frequency bands. Utilizing spectrogram analysis and signal processing, I achieved 95% accuracy, reducing clutter in 20 million sound samples. Expertise in PCA, t-SNE, and audio recognition enhanced results.
• 2. Additionally, I played a pivotal role in crafting a sophisticated geospatial localization map tailored for real-time tracking of underwater mammals. Leveraging my expertise in data analytics and machine learning, I meticulously engineered a comprehensive mapping tool that seamlessly integrated data from acoustic and GPS sensors. By employing advanced statistical analysis techniques and machine learning algorithms, I ensured the delivery of precise monitoring capabilities essential for extracting advanced research insights. This collaborative effort with esteemed researchers from MIT not only showcased my proficiency in data science methodologies but also demonstrated my ability to translate complex technical concepts into practical, impactful solutions that drive scientific exploration forward.
• 3. Assisted in redesigning an acoustic array for underwater mammal sound data analysis, implementing state-of-the-art machine learning solution for clustering species, and achieving a silhouette coefficient of 0.983. Developed a real-time Streamlit GUI for multi-processing analytics on 200M samples for 4k–500k frequency bands. Collaborated closely with researchers from MIT for data acquisition at the Gulf of Mexico expedition. This resulted in securing $4 million in funding for building a v2 with 6 fold subsection array.

Sep 2021 – Aug 2022 · 11 mos · Warsaw, Mazowieckie, Poland

• Mentored teams in the "KaggleDays x Z by HP Global Championship," showcasing leadership and decision-making skills within an agile team environment. This mentorship led to a notable 30% increase in participant engagement, translating into enhanced collaboration and knowledge sharing within the data science community. Additionally, by orchestrating 13 Kaggle competitions, we expanded the reach and impact of our initiatives, further solidifying LogicAI's position as a leader in data science innovation. Through these initiatives, we not only fostered a culture of continuous learning and improvement but also established valuable connections within the industry, paving the way for future collaborations and advancements in data science research.

Sep 2020 – Sep 2021 · 1 yr · Bengaluru, Karnataka, India

• 1. Engineered a dynamic, stateless retraining strategy, enhancing model adaptability across real-time transportation datasets. This strategy not only boosted model performance but also optimized the MLOps architecture, resulting in an estimated 20% reduction in operational costs. By building a robust data warehouse catalog, we streamlined data access and management, further amplifying cost savings and enabling scalability for future growth.
• 2. Collaborated closely with backend engineers to develop an ML platform for training and inference, integrating CI/CD pipelines for seamless deployment. Through the optimization of transport data governance using Databricks, we achieved a notable 20% reduction in congestion, translating into substantial cost savings for transportation agencies. Additionally, by enhancing MLOps efficiency by 15% and accelerating the ITMS forecasting pipeline by 1.5 times, we effectively expanded the project scope to handle larger datasets and more complex forecasting scenarios, ultimately driving additional value for stakeholders.
• 3. Spearheaded the development of an open-source ITMS forecasting data ingestion pipeline, processing 900k data points per minute from 5 smart cities. This initiative resulted in a remarkable 1.6x increase in real-time throughput, enabling city administrators to make more informed decisions with timely insights. The cost savings associated with this increased efficiency amounted to an estimated $X annually, demonstrating the tangible impact of our work on the bottom line.

Aug 2020 – Oct 2023 · 3 yrs 2 mos · Denver, Colorado, United States · Remote

• Spearheading as one of the 16 Global Data Science Ambassadors for HP, I strategically integrate cutting-edge hardware and software solutions to drive transformative outcomes in data science workflows. Through my extensive experience and expertise, I delved into collaborations with industry leaders like NVIDIA on building cutting-edge GPU technology that can benefit data scientists and researchers, optimizing data processing tasks with advanced hardware accelerators, and ensuring peak performance in data-driven initiatives.
• As a distinguished member of an elite cohort of Global Data Science Ambassadors representing Z by HP, I assumed a pivotal role in shaping thought leadership and fostering innovation on a global scale. By delivering captivating talks and presentations at prestigious events such as the GTC AI conference and ODSC West, I not only showcased the latest advancements in AI and data science but also facilitated meaningful collaborations and knowledge exchange among peers and industry stakeholders.
• Moreover, my passion for nurturing the next generation of data scientists extended beyond professional boundaries. I took pride in initiating various community events and mentorship programs, driven by a heartfelt commitment to empowering aspiring talent worldwide. Within the dynamic Kaggle community, I championed initiatives that offered students invaluable opportunities to learn directly from industry veterans. These efforts weren't just about imparting technical skills; they were about fostering a supportive environment where individuals could thrive, grow, and ultimately contribute meaningfully to the ever-expanding field of data science. My active involvement in mentoring and guiding budding data scientists not only enriched their learning experiences but also left a lasting impact on me, reinforcing the importance of mentorship and collaboration in shaping future leaders in the industry.

Jun 2020 – Sep 2020 · 3 mos

• Working as one of the core members of the founding team of AI Research and Development at Illumnus. Developing a smart Ai powered ecosystems by leveraging the power of Machine Learning and Deep Learning in our products focusing on collaborative based learning integrating institutions across various nations.

Apr 2020 – Apr 2021 · 1 yr · India · Remote

• Author at TDS, writing articles on Statistics, Machine Learning, and Data Science with around 10 -15k montly active readers. Link: https://blurred-machine.medium.com

May 2019 – Nov 2019 · 6 mos · Jaipur Area, India

• Illumnus is building a SaaS-based cloud ecosystem for educational institutes on the Android platform for Real-time Collaboration. Implemented Search Engine, Optimized Push Notifications using GCM & Firebase, Assignment Creating, Submission & Grading, Optical Bar code Detector, Performance Data Analysis & Administration Visualizations. Built a Recommendation System for teachers to get content sharing recommendations for students based on their teaching pattern.

May 2018 – Jul 2018 · 2 mos · Noida Area, India · On-site

• At Magicbricks I have worked as an Android Developer Intern mentored by Mr. Dhiraj Kumar. Contributed to the Text Recognition Classified in Magicbricks Android Application. Major work fields include Optical Character Recognition API, Location Tracking, Map Clustering, AutoComplete API, Canvas Animations, Battery Usage Optimization, Android Permissions library, Deep Link and Firebase.