Jul 2021 – Sep 2023 · 2 yrs 2 mos

• Excelled in requirements gathering, and coordinating with various stakeholders like CCEs, product managers, documentation writers, customer support, resulting timely delivery.
• Effective collaboration, requirement gathering, delivery of production-ready code along with suitable tests within strict timelines.
• Demonstrated agility in adapting to evolving product needs and consistently ramping up on new features to drive success.
• Full Stack Developer:
• Code highly scalable software components based on the latest version of LWC(React like), TypeScript, JavaScript, Java, Spring.
• Build UI in EBV feature for showing massive user data
• Build complex web components using TypeScript and related technologies.
• Build highly reusable and shared components.
• Translate designs from low-fit prototypes provided by the UI/UX
• Contribute to the UI style guides and a UI component library, creating reusable components
• Write application interface codes using JavaScript following LWC (react like ) workflows.
• Design UI Screens that are responsive across all devices.
• Build and maintain our backend services powering the core of our software platform
• Work with highly performant team of 7 members to maintain high uptime and reliability of our core backend and APIs
• Work as a part of Agile Team building backend system to scale and perform as per the need.
• Build and maintain complex backend services which are highly scalable and highly available.
• Use the best practices in the Industry to write reusable code.
• Skills: JavaScript, Java, SpringBoot, Node.js, LWC (UI) framework, Jest, Perforce, Git, Jenkins.

Jan 2021 – May 2021 · 4 mos · On-site

• During my tenure with the Platform R&D Team at SRI Noida, I had the opportunity to work on an exciting project involving slow motion video generation with synchronized background music. The focal point of this endeavor was a modal system capable of generating music that perfectly complemented the slow frames produced.
• Throughout the project, we explored and studied different approaches and considerations in video frame interpolation, including:
• 1. Optical Flow-Based Interpolation:
• We investigated methods that utilized optical flow estimation to generate intermediate frames by analyzing the motion patterns between consecutive frames.
• 2. Deep Learning-Based Interpolation:
• We delved into the application of deep learning techniques, such as convolutional neural networks (CNNs) and recurrent neural networks (RNNs), to learn the temporal dependencies and generate interpolated frames.
• 3. Depth-Aware Interpolation:
• We also explored techniques that took into account depth information to improve the accuracy of frame interpolation, leveraging depth maps or other depth estimation methods.
• By working on this project, I gained valuable insights into handling multi-modal datasets, as we had to combine video frames and corresponding music data to train and evaluate our model effectively.
• Overall, this project was a valuable learning experience that allowed me to delve into the fascinating intersection of computer vision, machine learning, and audio processing, while enhancing my expertise in handling multi-modal datasets

May 2020 – Jul 2020 · 2 mos

• 🌱As an attempt revolutionize the 🌿🪴agricultural industry and mitigate crop losses caused by plant diseases.
• ✨ Project Overview:
• In collaboration with a team of talented researchers and engineers, I developed an deep learning solution to detect plant diseases accurately and efficiently through leaves image. The project aimed to improve crop yield.
• 🔍 Key Features:
• 1️⃣ Dataset Creation: We curated a comprehensive dataset consisting of thousands of high-resolution images capturing various healthy and diseased plant samples. This dataset was carefully annotated by domain experts, providing invaluable ground truth for training our models.
• 2️⃣ Deep Learning Models: Leveraging state-of-the-art deep learning architectures, including Convolutional Neural Networks (CNNs), we trained a robust and highly accurate disease detection model.
• 3️⃣ Transfer Learning: To overcome the challenge of limited labeled data, we utilized transfer learning techniques. By leveraging pre-trained models such as ResNet, VGGNet, we achieved remarkable results even with a relatively small dataset. This approach significantly reduced the time and resources required for training while maintaining high performance.
• 🌿 Impact and Results:
• Through extensive testing and validation, our Plant Disease Detection project demonstrated exceptional performance, achieving an accuracy rate of over 95% in identifying various plant diseases.

May 2019 – May 2019 · 0 mo · IIT Patna

• understanding the emotions and sentiments expressed in emails..
• ✨ Project Overview:
• The project aimed to analyze and classify the sentiment of emails accurately and efficiently.
• 🔍 Key Features:
• 1️⃣ Dataset Annotations
• The dataset encompassed a wide range of topics, sentiments, and language styles, providing a representative sample of real-world email communications. Label each sentiment on scale 1 to 5.
• 2️⃣ Preprocessing and Text Representation:
• I explored text preprocessing techniques, such as tokenization, stop-word removal, and stemming, to transform the raw email text into a format suitable for analysis. Additionally, we leveraged techniques like TF-IDF (Term Frequency-Inverse Document Frequency) and word embeddings (e.g., Word2Vec, GloVe) to represent text semantically and capture contextual information effectively.
• 3️⃣ Machine Learning Models: Using supervised learning algorithms, such as Support Vector Machines (SVM), Naive Bayes, and Random Forests, we trained a sentiment classification model.
• 4️⃣ Model Evaluation and Fine-tuning: We employed various evaluation metrics, including accuracy, precision, recall, and F1-score, to assess the performance of our sentiment analysis model

Oct 2018 – Apr 2019 · 6 mos · National Institute of Technology, Patna

Oct 2018 – Apr 2019 · 6 mos · Patna Area, India

• "Earth wouldn’t be better off without humans, despite our faults..." ;) that's why we are developing new human assets for a better tomorrow here @sankalp.nitpatna ...