"Data is the new oil, and the ability to sift through it effectively is not just a skill it's a necessity," says Natasha Mittal, a name increasingly associated with innovation in software technology. As we navigate the complex corridors of tech, where code and algorithms are the building blocks, Mittal is a guiding light with her groundbreaking research and practical solutions.
The Professional Journey of Natasha Mittal
Natasha Mittal is far from a typical software engineer. With over seven years of hands-on experience developing complex distributed systems, she currently spearheads the Search Experience at DoorDash New Verticals. This division focuses on convenience and grocery shopping. Before DoorDash, she was a pivotal member of the Generic Pollers team at Amazon AWS. There, she created a platform that helped major clients like McDonald's and OkCupid integrate with AWS Lambda, generating a GMV of over $100 million within the first month.
Prior to Amazon, Mittal honed her skills at WizeCommerce as a full-stack engineer. She was responsible for designing and implementing complex products and features, which laid the groundwork for her future roles at Amazon and DoorDash.
Moreover, the software expert has spearheaded innovations at DoorDash, revolutionizing user experience and search functionality, which has driven sales and elevated user satisfaction. Her groundbreaking work includes the award-winning SNAP project and the development of a custom relevance pipeline, setting new standards in search quality and positioning DoorDash competitively against rivals like Uber Eats and Instacart.
Academic Contributions: Bridging Theory and Practice
Building on her industry experience, Mittal's influence extends to the academic world. Her research on steganalysis, a field focused on detecting hidden data in digital files, has garnered attention. Her paper, "FS-SDS: Feature Selection for JPEG Steganalysis using Stochastic Diffusion Search," was accepted by IEEE and presented at the Systems, Man, and Cybernetics (SMC) conference in 2014.
The paper introduces a novel algorithm that employs Stochastic Diffusion Search (SDS) to enhance the efficiency and accuracy of steganalysis. "Efficiency is as crucial as finding the hidden data," the software engineer observes, emphasizing the research's real-world applicability in cybersecurity and digital forensics.
Innovations in Edge-Cloud Computing
Besides her work in steganalysis, Mittal has also ventured into edge-cloud computing. Her research on Cooperative Log Structured Merge Trees (CooLSM) aims to tackle some of the most urgent issues in data storage and indexing. "The traditional structure of LSM trees has been a bottleneck for scalability and flexibility. CooLSM deconstructs this to create more manageable components," Mittal elaborates.
This innovative approach has practical implications. For instance, the research addresses the challenges of write performance in edge-cloud environments, a significant issue in today's distributed systems. "By breaking down LSM trees, we can optimize resource allocation and data management, particularly for real-time applications like Vehicle-to-Everything (V2X) tasks," she adds.
The development of CooLSM was motivated by the limitations observed in the existing paradigms of cloud and edge computing. Conventional cloud computing, dependent on remote data centers, introduced substantial restrictions for applications requiring real-time guarantees. On the other hand, relying solely on edge computing was deemed unsuitable due to the restricted hardware capacity of edge devices. Mittal conceptualized CooLSM as a bridge between edge and cloud resources, aspiring to transform edge-cloud data management systems. This system is meticulously designed to address the challenges of managing storage and access to data where processing and storage are dispersed across edge and cloud nodes. It caters to the urgent need for rapid data ingestion crucial for edge and IoT applications.
Real-world Applicability and Performance Gains
CooLSM, a revolutionary distributed data indexing system, holds significant applications in IoT, especially in innovations related to Smart City and Industry 4.0. For instance, CooLSM enables rapid communication between vehicles and surrounding entities in smart city traffic applications, managing instantaneous data ingestion and access at the edge. This technology is pivotal for allowing vehicles to interact with pedestrians, traffic signals, and other vehicles through Vehicle-to-Everything (V2E) technology. While CooLSM manages immediate communication and simpler tasks at the edge, intricate analyses, such as evaluating traffic patterns, are conducted in the cloud, ensuring optimal resource utilization and efficiency. Furthermore, the adaptability of CooLSM extends to Virtual/Augmented Reality-based mobile games and social networks, demonstrating its wide-ranging applicability.
Mittal's CooLSM research was rigorously tested in the context of smart traffic systems, demonstrating its feasibility in real-world scenarios. The results were promising: CooLSM achieved low latency, crucial for real-time applications, and showed potential for high scalability.
"This could revolutionize applications like smart cities and Industry 4.0, where real-time processing at the edge and complex computations in the cloud are essential," she notes.
Industry Implications and Future-Proof Design
The implications of Mittal's research are far-reaching, especially for burgeoning fields like IoT and edge computing. "The low latency and scalability of CooLSM make it a game-changer for these industries. It could also influence the design of future database technologies that require high performance and scalability," she explains.
Moreover, the paper explores opportunities for scalable and elastic behavior of individual components, indicating a future-proof design. "CooLSM is not just an academic exercise; it's a tangible solution to some of the industry's most pressing challenges," she concludes.
Mittal's exploration into the creation of CooLSM started with her profound research in database indexing techniques, focusing specifically on Log Structured Merge (LSM) trees. Fascinated by the capabilities of LSM trees, she immersed herself in many scholarly papers, journals, and conferences to gain an all-encompassing understanding of the subject. Her pursuit of knowledge culminated in creating CooLSM, an innovative indexing system designed to resolve the challenges associated with edge-cloud indexing. The software expert realized a functional prototype of CooLSM and articulated her findings in a research paper, achieving acknowledgment at the prestigious IEEE ICDE (International Conference on Data Engineering) 2021 conference.
A Skeptic's Contrasting View
While Mittal's work has received widespread acclaim, it's worth noting that some critics remain skeptical. "Her research is inventive and original, but the true test is its practical application," says another expert in the field of software engineering.
This skepticism serves as a reminder that innovation must meet practical application. While Mittal's research offers ingenious insights, their real-world applications will determine their true value—a challenge Mittal seems well-equipped to meet.
An Expert's Touch
"In the future, the software industry will focus less on data accumulation and more on effective data interpretation," says Natasha Mittal, summarizing her vision for a world where data and human ingenuity merge to create something extraordinary. Her impressive technical skills, resourceful thinking, and commitment to social responsibility make her a transformative figure in the software industry.
With her research and practical solutions spanning user experience to edge-cloud computing, Mittal is shaping her own future and charting a new course for the entire industry.
