At Balbix, we heavily leverage our expertise in ML/AI, probabilistic modeling, signal processing, neuroscience to develop production grade scalable algorithms for predictive security breach analysis. We enable our customers to proactively and comprehensively secure their enterprise networks before they get breached.Key achievements:- First and only ML engineer from idea to launch.- Architected the entire ML infrastructure and production pipeline.- Developed, deployed and scaled classical machine learning and deep learning algorithms.- Leveraged signal processing background to develop bespoke ML algorithms employing Probabilistic Graphical Models, Bayesian networks, HMMs, Hierarchical logistic regression.

– Drive enhancements in PHY layer performance, algorithm complexity and power consumption for Qualcomm’s next-generation modem chipset.– Develop algorithms to enhance performance in large delay-spread channels without increasing overall complexity.– Innovate solutions for minimizing the search and signal acquisition performance in a deployment with multiple technologies occupying the same frequency band.– Responsible for maintaining a complex modeling, simulation and performance verification framework.

– Led design of major core functionalities to counter RF interference issues for Qualcomm’s industry-leading first fully integrated Multi-SIM modem chipset (MSM8x26).– Developed novel baseband processing solutions to minimize standby power consumption which gave Qualcomm’s products a competitive edge. These features are used in most of the leading smartphones that have Qualcomms modem chipsets.– Led global coordination efforts across system design, software, hardware, firmware and test teams.– Actively involved in hardware bring-up activities, functional/performance testing and debugging.

– Designed differentiating PHY/MAC interference and mobility management algorithms for self-organizing femtocell networks . These algorithms facilitate unplanned and autonomous deployment of low-power, plug-n-play cellular base-stations (femtocells) and improve overall system capacity and performance.– Key member of the team that designed and commercialized self-configuration algorithms for Qualcomm’s first 3G femtocell chipset.– Multiple related granted patents and publications.

– Designed and prototyped key aspects of femtocell interference management and performance improvement such as self-configuration, transmit diversity, wireless channel characterization and timing & frequency synchronization.– Made significant contributions to the standardization of the femtocell OA&M interface in the 3rd Generation Partnership Project 2 (3GPP2).

– Analyzed and implemented various interference cancellation schemes in C++ for 4G LTE MIMO.– Did detailed complexity/performance study.

– Implemented low-complexity turbo equalization and channel estimation schemes on TI TMS320C6201 32-bit DSP in C.– Results published in international conferences.