I’m creating new ML and GenAI solutions for client-facing use cases with technologies developed by IBM Research.

(FORMERLY PRIDATEX) The industry need I solved with tomtA.ai is the democratization of private data held up in our clients’ data silos to facilitate collaboration on building more effective AI systems.TomtA.ai is at the forefront of creating an innovative Differential privacy-based data generation platform, ensuring organizations can confidently anonymize real data without losing its intrinsic value.During my journey, I've been instrumental in a multitude of aspects encompassing the invention of the method, design, development, and deployment of our solutions. Key accomplishments I'd like to share:Innovation in Algorithms: Spearheaded the ideation/research to extend Pridatex novel Differential Privacy algorithm with Generative AI algorithms in Python, Scikit-Learn; Improved privacy by 30% and accuracy by 10%Data Handling: Undertook extensive dataset anonymization projects, ensuring robust and secure data processing and management.Architectural Mastery: Conceived and executed a full-stack platform using the MERN framework. Recognizing evolving client needs, I transitioned our operations to a Serverless architecture, enabling unparalleled scalability and flexibility. Deployed algorithms with Docker onto partner marketplaces to advertise to 2x amount of potential clients.Cloud Integration: Seamlessly facilitated operations across prominent cloud platforms like AWS while adeptly handling sensitive on-premise data with a Hybrid Cloud Architecture.Collaborative Ventures: Led 5 successful POCs with distinguished US and European clients. This synergy allowed us to gather invaluable feedback and refine our offerings, setting new benchmarks for excellence in the industry.My time at TomtA.ai has been a mosaic of relentless innovation, collaborative growth, and unwavering dedication to customer satisfaction. I remain deeply committed to pushing the boundaries of machine learning and data privacy, and I eagerly anticipate future challenges and milestones.

At Pridatex, I helped advance the current state of AI by pioneering differential privacy solutions that strike a balance between ensuring data privacy and retaining its inherent value. It improved rare disease detection on anonymized datasets by 70% compared to synthetic data and privacy by at least 80% as we had a mathematically backed method of guaranteeing it.In my leadership capacity, some of the achievements I am most proud of include:Technical Innovation: Driving the research and development of a patented Differential Privacy data product. Utilizing a diverse toolkit with various OOP languages along with Python and Scikit-learn, our team successfully built a product that set new standards in data privacy.Market Strategy: I directed our strategic alignment with prevalent market and privacy standards. It's noteworthy that our commitment to excellence paved our acceptance into the esteemed Plug and Play Insurtech Batch 8.Strategic Partnerships: Instrumental in forging a significant partnership with BigID, we seamlessly integrated our solution with their renowned Data Discovery Platform, further enhancing our product's market standing.Platform Versatility: Overseeing the development across diverse platforms such as Linux, Microsoft, and Mac ensured that our product was versatile and user-centric. Its deployment on AWS EC2 amplified its accessibility and scalability.Customized Solutions: Recognizing the diverse needs of our clientele, I ensured we tailored interfaces for a wide range of SQL databases like IBM DB2, Oracle, MariaDB, and MySQL, catering to specific client requirements.Leading Pridatex was a journey of fostering innovation, ensuring market alignment, and relentlessly pursuing excellence. These experiences have fortified my commitment to leadership in tech, and I look forward to the next chapters of my professional journey with the same zeal and enthusiasm.

From September 2017 to May 2018, I had the incredible opportunity to delve into the world of database privacy under the guidance of Prof. Christopher W. Clifton. Our research focused on enhancing individual privacy in hierarchical databases like the ones used by the US Census Bureau.Merging Privacy Methods: Our primary objective was to combine the principles of Differential Privacy and k-Anonymity. This amalgamation aimed to bolster individual privacy protections for categories represented within hierarchical databases, such as US Census Data.Sparse Vector Technique Enhancement: Beyond the merger of established methods, we further investigated how to improve the Sparse Vector Technique. By strategically allocating privacy budgets to specific ranges and aggregate categories within a hierarchical database, we were able to save privacy budget without compromising data accuracy.Funding and Impact: Supported by an NSF Research Grant, our work not only contributed to the theoretical understanding of database privacy but also had practical implications for the management and release of sensitive data by government agencies.Lasting Learning: The research experience with Prof. Clifton has shaped my perspective on the critical balance between data utility and privacy. It was a rewarding exploration that allowed me to contribute to an essential and ever-evolving field, and it continues to inform my approach to data privacy and security.

Beginning in October 2016 and culminating in a publication in February 2018, I undertook a groundbreaking research project under the guidance of Prof. Emiritus B.J. Lucier. This innovative work led to the development of a novel polynomial regression method known as Finite Difference Regression (FDR), which was subsequently published in the Open Journal of Statistics.Innovation in Polynomial Regression: Finite Difference Regression challenges conventional approaches. Unlike traditional methods, it utilizes the t-test combined with finite differences, offering a more sensitive and objective measure of the order of the best-fitting polynomial for uniformly sampled sequences of noisy data points.Efficient Coefficient Estimation: By reemploying the finite differences used in determining the order, I developed a technique that provides excellent estimates of the coefficients of the best-fitting polynomial. Not only are these coefficients unbiased and consistent, but the asymptotic properties of the fit also improve with increasing polynomial degrees.Publication and Recognition: My original findings were carefully reviewed and published in the Open Journal of Statistics, a testament to the quality and impact of this research. The publication provides a comprehensive look at the methodological innovations and practical applications of FDR in statistical analysis.Lasting Contribution: Finite Difference Regression represents a significant advancement in the field of statistics. By introducing a more precise and nuanced approach to polynomial regression, this research opens new avenues for data analysis and modeling. The principles and techniques developed continue to influence my work and highlight the ever-evolving nature of statistical methodology.

From September to December 2017, under the mentorship of Prof. B. Dunsmore, I embarked on an exciting journey to innovate within the Agile framework. My team and I conceptualized and developed a new approach called "Lean Scrum" aimed at enhancing efficiency and reducing waste in the software development process, particularly within an entrepreneurial or product development setting.Innovation in Agile Framework: Lean Scrum creatively fuses the methodologies of Lean Startup and Scrum, forming a more dynamic Agile process. This unique combination improves team learning during product development and addresses some problematic issues within traditional Scrum.Lean Scrum Principles: By integrating novel methods from Lean Startup, Lean Scrum emphasizes reducing time and resource waste, enhancing customer traction, and fostering a more adaptive development process. It represents a paradigm shift in how software can be developed, especially for entrepreneurs looking to be more nimble and responsive to market demands.Presentation and Recognition: Our original findings were proudly presented at the 2018 Purdue Undergraduate Research Conference. This opportunity allowed us to showcase our work to a broader audience and receive valuable feedback from peers and experts in the field.Lasting Contribution: The development of Lean Scrum wasn't just an academic exercise; it was a real-world contribution to the continually evolving field of software engineering. It reflects a passion for innovation, collaboration, and the pursuit of efficiency. I continue to draw on the principles and practices of Lean Scrum in my ongoing work, recognizing its potential to transform the way software products are created and brought to market.

As Founder and CEO at Visorware from June 2016 to March 2018, I was driven by a passion for enhancing driver safety through innovative image processing and computer vision technology.Problem and Solution: Current technology in heads-up displays (HUDs) primarily adjusts projections' luminosity, often leading to visibility issues (e.g., a bright green projection on a dark green background). Recognizing this challenge, I developed a patented Image Processing algorithm during my college years that uses color-change to make HUDs more visible, less distracting, and significantly safer for drivers to reduce visibility related accidents by 30%.Innovative Solution: The software around our Image Processing Algorithm mitigated the adverse effects often seen with brightness adjustments on images projected onto visors of HUDs/AR displays. Written in C/C++, and utilizing OpenCV for Image Segmentation to average our algorithm’s output over each connected component, we applied this solution not only to HUDs but also to augmented reality (AR) products, especially in workplaces with hazardous conditions. Industry Recognition: Our technology caught the attention of both large clients such as Continental and small startups working in HUDs and AR. Various articles, including one on PCMag, have also cited visibility issues in HUDs, further validating the importance of our approach.Market Impact: At Visorware, we were not just about technology; we were about ensuring safety and enhancing focus on the road. Our software represented a breakthrough in the industry, providing a tangible solution to a longstanding problem.Vision for the Future: The traction we've gained speaks to a broader application of our technology across diverse industries, reflecting the universal importance of visibility and safety. I'm thrilled to have contributed to a product that continues to have a positive and lasting impact.

During my internship at DCG Systems Inc. (now FEI) from May to August 2015, I was immersed in the cutting-edge field of microchip defect detection, contributing both to core tasks and innovative side projects.Image Processing Expertise: As an Image Processing Algorithms Intern, I co-developed camera control and image processing software to detect defects in microchips. Working with C# and MATLAB, I played a critical role in refining and implementing algorithms that enhanced the accuracy and efficiency of our microchip inspection systems.Innovative User Interface: Outside of my primary responsibilities, I also devoted my free time to creating a smarter Graphical User Interface (GUI). This GUI was designed to predict when a user had stopped providing input, employing statistical analysis to achieve an intuitive and responsive user experience.Invention Recognition: My work at DCG Systems was not only about learning but also about innovation. I disclosed my first Invention Alert Form and successfully licensed this invention to the company. It was a proud moment that underscored the value of creativity, curiosity, and technical skill.Lasting Impact: My time at DCG Systems was a rich blend of technical challenges, collaborative work, and individual innovation. I left not only with new skills but with the satisfaction of having contributed to a product that plays a vital role in maintaining the quality of microchips. The experience continues to influence my approach to technology and innovation.