Rudra Sikri

Hi, I am Rudra 👋

Empowering Ideas Through Technology

About Me

grid-1

Hi, I'm Rudra

With 1+ year of hands-on experience in software engineering, I have had the opportunity to work on a variety of challenging projects that have helped me develop a strong foundation in software architecture, backend development, and cloud infrastructure.

grid-2

Tech Stack

I specialize in Django, React.js, Node.js, Python, JavaScript, Docker, and have experience with a wide range of tools and technologies.

I work remotely across most timezones.

I'm based in India, with a flexible schedule to work remotely across the globe.

grid-4

My Passion for Coding

I love solving problems and building products that make a difference. I'm always eager to learn new technologies and take on new challenges.

grid-5

Contact Me

copy

rudrasikri12@gmail.com

My Selected Work

spotlight
logo

Prediction of Warfarin Dosage using MAB

Summary:

Developed a Warfarin dosage prediction model using cutting-edge reinforcement learning and achieved an impressive 68% accuracy, outperforming traditional methods by 18%. Utilized a novel MAB algorithm in the model, allowing for individual patient optimization and continuous learning from experience.

Description:-

Problem Statement - Warfarin is a commonly prescribed anticoagulant medication used to prevent blood clots. However, the dosage of Warfarin is highly dependent on the patient's genetic makeup, which can vary significantly from person to person.

As a result, the dosage of Warfarin must be carefully monitored and adjusted to ensure that the patient is receiving the correct amount of medication.

However, this process is time-consuming and requires frequent blood tests to ensure that the patient's anticoagulation levels are within the desired range.

As a result, patients often experience frequent and unpredictable fluctuations in their Warfarin dosage, which can lead to serious health complications.

 

Approach - Our model was trained on a dataset of 5,700 patients, each of whom had their Warfarin dosage adjusted according to their INR (International Normalized Ratio) levels.

We used a Linear UCB Contextual MAB algorithm to select the optimal dosage for each patient, allowing for individual patient optimization and continuous learning from experience.

We then used the model to predict the optimal dosage for new patients, We compared the model's performance to that of a traditional model, such as fixed dose, and Clinical Dosage Algorithm.

We found that our model achieved an impressive 68.09% accuracy, outperforming the traditional models.


github

My Work Experience

Software Engineer

CRUVSept 2024 - Present

Remote

  • Developed an automation tool in Python to streamline the design-to-development process by converting Figma designs into Framer, reducing manual work and increasing design efficiency.

Software Engineer Intern

CRUVMay 2023 - Aug 2024

Remote

  • Engineered and led backend architecture development for Craver (an ONDC-based food delivery app), collaborating closely with experienced industry professionals.
  • Implemented microservices distributed systems architecture, REST APIs, cloud infrastructure and optimized algorithms to handle high concurrency, scalability, fault tolerance and seamless system performance.
  • Integrated Payment gateway into backend system to ensure secure, and efficient transactions.
  • Designed and deployed pub/sub architecture, utilizing message queues to efficiently ingest and manage burst payloads of up to 30K requests per minute, enhancing system responsiveness and scalability.
  • Optimized backend performance by reducing 50% processing time and 30% memory usage through optimum algorithm improvements and proactive bottleneck identification. Achieved a 20% reduction in API response time.

Data Science and Machine Learning Intern

Gilbert Research CenterApr 2023 - Jun 2023

Remote

  • Developed a deep understanding of mathematical concepts such as linear algebra, calculus, and probability theory, which are foundational for data science and analytical skills.
  • Implemented feature engineering techniques like principal component analysis (PCA) and feature scaling to prepare datasets for a machine learning model, leading to a 10% increase in model accuracy, showcasing advanced data analysis abilities.
  • Leveraged data preprocessing techniques, including data cleaning and missing value imputation, to improve data quality, resulting in a 5% reduction in misclassification rates, further solidifying data science proficiency.

Product Development Intern

RadicalXOct 2022 - Nov 2022

Remote

  • Facilitated overhaul and redesign of an interface for clients which increased the efficiency and ease of use.
  • Worked on an interactive platform based on React.
  • Collaborated with a diverse group of creators and developers from all over the world.

Let's talk

Have a question or want to work together? Feel free to reach out!