Apr 2025 – May 2025 · 1 mo

Apr 2024 – May 2025 · 1 yr 1 mo

• ● Engineered a streamlined authentication system with modal-based login accessible site-wide, reducing login time by 78% and eliminating
• redundant redirects, enhancing user experience while improving codebase maintainability and simplifying overall web login architecture.
• ● Single handly developed a microsite for SVIP under tight deadlines; leveraged this foundation to develop CANVAS, an enterprise-level
• hackathon platform. Demonstrated code reusability by adapting the initial template, resulting in 3-4 days of saved development effort.
• ● Engineered a Creator Dashboard, the dashboard’s key features included secure creator authorization, an intuitive post-creation page, and
• a versatile analytics dashboard, providing creators with actionable insights to optimize their content strategies and improve audience reach.
• ● Spearheaded Xclusives webpage; integrated Strapi CMS, automating event management and registration.This initiative reduced Ops team
• manpower by 66% (from three to one), while enhancing operational efficiency and improving candidate engagement and website traffic.
• ● Led a 3-person team in developing CANVAS, a comprehensive hackathon platform. Architected candidate-facing features (registration,
• team formation, multi-round assessments) and enterprise tools (demographic analytics, submission evaluation), driving company's expansion into the hackathon market.

Jun 2022 – Mar 2024 · 1 yr 9 mos

• 1. Developed & automated the Sales-Assisted Buy Coin Flow, significantly enhancing user experience for coin purchases and payments, the flow improved sales team efficiency by streamlining order management and payment tracking, leading to better operational effectiveness.
• 2. Contributed to cross-functional team success in delivering a user-friendly Enterprise Client dashboard with seamless UI integration.
• 3. Refactored the Monetization codebase to improve readability and maintainability, aligning with business and product needs. Enhanced page
• performance by 5% through Lazy Loading & minimizing unnecessary re-renders, delivering a more efficient & responsive user experience.

Apr 2021 – Aug 2021 · 4 mos · Berlin, Germany · Remote

• Upgraded the AllenNLP-based UDIFY model for universal part-of-speech tagging, dependency parsing, and morphological features.
• Developed ML models to extract semantics from the Universal Dependency Treebank dataset, achieving an accuracy rate of 97.5%.
• Analysed and labeled English Universal Dependency Treebank data with location and time information to empower semantic extraction.

Jul 2020 – Aug 2020 · 1 mo · New Jersey, United States · Remote

• Conducted research and implemented various Adversarial Attacks on Machine Learning models.
• Investigated and applied Empirical Robustness and Loss Sensitivity scores for Convolutional Neural Networks, as well as Verified Error and Average Robustness scores for Tree-based classifiers to quantify the impact of Adversarial Attacks.
• Employed preprocessing mitigations and defensive techniques, including Feature Squeezing, Gaussian Data Augmentation, and Spatial Smoothing, to protect Machine Learning models trained on both structured and unstructured datasets.

Apr 2020 – May 2020 · 1 mo · Bengaluru, Karnataka, India · Remote

• Conducted thorough analysis and preprocessing of a raw dataset comprising cancer patient records, focusing on tumor size and the impact of prescribed treatment drugs.
• Employed Google Colab for Transfer Learning and Fine-tuning, utilizing a pre-trained ResNet50 model on the NIH ChestX-ray dataset.

Aug 2019 – Apr 2020 · 8 mos · IIT Kharagapur

Aug 2018 – Apr 2019 · 8 mos · IIT kharagpur