About
"AuroiTech" specializes in developing eye care software solutions, including EMR, HIS, Optical Management, Lab Information, and Material Management systems. Their cloud-based EMR and HIS systems are used in over 120 hospitals in India and 150+ eye hospitals internationally. With three decades of collaboration with Aravind Eye Hospitals, AuroiTech has developed high-quality, cost-effective software products on par with global standards.
The company, in partnership with Google and others, has created AI-enabled ophthalmology solutions. AuroiTech aims to reduce blindness and improve lives by connecting IT work to this mission. The company has development centers in Madurai and Chennai, focusing on enhancing EMR solutions, and plans to build IoT and AI/ML solutions for healthcare automation.
Rounds
Hackathon
The process starts with sharing the dataset during the hackathon, followed by assigning use cases for the participants to work on. Once completed, they submit their models on Google Form. Finally, the submissions are evaluated and screened based on their scores. Selected candidates will receive further communication, and the results will be posted on the website.
Solution Discussion
The process begins with participants preparing a PPT to showcase their solutions. They then present their work, followed by a discussion where questions about their approach are addressed. Finally, a code walkthrough is conducted to review their implementation.
Technical Interview
The process involves a 1-hour technical interview focusing on machine learning and deep learning concepts. During the session, the candidate's previous projects are discussed to assess their experience and problem-solving approach.
Job Roles
3 to 5 years
Responsibilities :- Drive the development and deployment of machine learning models to solve complex business problems, working closely with cross-functional teams.
- Guide junior data scientists and collaborate with business stakeholders to align data strategies with organizational goals.
- Continuously improve model performance through tuning, feature engineering, and evaluation using industry-leading techniques.
- Research and implement the latest advancements in AI, ML, and data science to keep solutions innovative and competitive.
- Translate data insights into actionable strategies, ensuring that models drive tangible outcomes and support key business decisions.
0 to 3 years
Responsibilities :- Work with data to build, test, and deploy machine learning models to address business challenges.
- Partner with senior data scientists and cross-functional teams to understand business needs and contribute to data-driven solutions.
- Clean, preprocess, and transform data to ensure it’s ready for analysis and model building.
- Evaluate model performance, experiment with different techniques, and contribute to improving model accuracy and efficiency.
- Continuously learn new tools, technologies, and methodologies in AI/ML, staying up-to-date with industry trends to enhance your skills.
Themes
JOB : Sr. Data Scientist
This challenge invites participants to design and implement a Retrieval-Augmented Generation (RAG) system tailored for healthcare applications. The goal is to develop a system that can understand user queries related to patient history, retrieve the most relevant information from a database, and generate accurate, concise, and contextually appropriate responses.
Problem Statement :In healthcare, patient history plays a vital role in diagnosis and treatment. Medical practitioners and administrative staff often need quick access to relevant data from vast databases. However, retrieving precise information based on a query can be challenging due to the complexity and variability of medical data.
- Understands user queries : Handle both structured (e.g., predefined templates) and unstructured (free text) queries.
- Retrieves relevant data : Efficiently query a database containing patient history to extract pertinent information.
- Generates natural language responses : Summarize or elaborate on the retrieved data to address the query.
- Scalable to handle large databases.
- Robust in dealing with noisy or incomplete data.
- Privacy-compliant to ensure sensitive patient data is protected.
- Query Understanding : Implement NLP techniques to parse and interpret queries, supporting multi-turn interactions.
- Contextual Retrieval : Use techniques like vector search or keyword-based filtering to retrieve relevant patient records.
- Answer Generation : Leverage models (e.g., GPT or other LLMs) to generate human-readable summaries or recommendations based on retrieved data.
This challenge calls upon participants to design and implement a robust Image Detection System tailored for healthcare applications. The goal is to create a system that can analyze medical images, accurately detect anomalies or specific features, and provide actionable insights to aid healthcare professionals in diagnosis and treatment planning.
Problem Statement :Medical imaging, such as X-rays, MRIs, CT scans, and histopathological images, plays a critical role in healthcare. Accurate interpretation of these images is vital but time-consuming for medical professionals. Participants are tasked with building an AI-powered system capable of automating the detection of medical conditions from images while ensuring high accuracy and reliability.
- Detect anomalies or features : Identify regions of interest in medical images.
- Classify conditions : Distinguish between normal and abnormal findings or between different conditions.
- Localize findings : Provide bounding boxes or heatmaps for regions of interest where applicable.
- Accurate : High precision and recall rates to minimize false positives/negatives.
- Preprocessing : Handle noisy, low-resolution, or varied quality images through robust preprocessing pipelines.
- Model Design : Use state-of-the-art computer vision models (e.g., CNNs, Vision Transformers) fine-tuned for medical imaging.
JOB : Data Scientist
This challenge invites participants to design and implement a patient volume forecasting system that predicts the number of patients expected at a healthcare facility over a given time horizon. The system will aid hospitals, clinics, and healthcare administrators in resource planning, staff scheduling, and inventory management to improve operational efficiency and patient care.
Problem Statement :Accurate forecasting of patient volume is critical for healthcare facilities to manage their resources effectively. Underestimating patient volume can lead to long wait times and overworked staff, while overestimating it can result in wasted resources. Participants are tasked with building a system that leverages historical data and external factors to forecast patient inflow with high accuracy.
- Predicts patient volume for Short-term (daily or hourly forecasts for the next week).
- Predicts patient volume for Medium-term (weekly forecasting for the next month)
- Temporal patterns (e.g., weekly, monthly, or seasonal trends)
- External factors affecting patient inflow.
- Data Preprocessing : Handle missing, noisy, or outlier data effectively.
- Forecasting Model : Use advanced time series modeling techniques such as ARIMA, Prophet, LSTMs, or transformer-based models.