Oct 2023 – Feb 2024 · 4 mos · Bengaluru, Karnataka, India

Mar 2022 – Nov 2023 · 1 yr 8 mos · Bengaluru, Karnataka, India

• Working in design and development of Cloud Native Mobility Management Entity(CN-MME) as well as Azure Private 5G core (AP5GC), which is a highly scalable product with a highly available, fault tolerant microservice architecture , responsible for handling LTE and 5G network attach with millions of subscribers , which is deployed in AZURE Cloud.
• DNS server response caching in MME application:
• Handled this feature end-to-end on caching the DNS server response in local application pods by interacting with ETCD. while ensuring data consistency, scalability and fault tolerance. This decreased the latency and improved the perfomance.
• Accessing subscriber data from outside Kubernetes cluster:
• Worked on exposing ingress Gateway pod outside application boundary via Istio Virtual Service so that the network specific subscriber data can be accessed outside Kubernetes Cluster and eventually the data can be displayed in AP5GC Azure portal
• Design and Development of Service Assurance Server(SAS):
• SAS is a distributed logging system that traces logs between various microservices that are deployed in the overall application. It uses Jaeger backend which is an Open Source logging system
• Developed Log-analysis tool for analyzing call flows:
• Developed a Python script containing specific filters that would parse a given log and help analyze the code flow and markers

Jun 2021 – Mar 2022 · 9 mos

• Working in Snapdragon Neural Processing Engine(SNPE) team
• Inference and Benchmark of Neural network models: ->
• Added support for 50+ different neural network models, which included models for Image segmentation, Image classification, Object detection, IOT applications. Each model support included:
• Visualizing the input and output layers of the neural network models
• Generating framework (Tensorflow/Caffe/Caffe2/Onnx) output on CPU host and SNPE output on different processors(CPU,GPU, DSP) running in backend
• Implementing verification algorithms like Cosine Similarity,Rtol Atol to compare framework and SNPE generated outputs
• Generating total inference time for Benchmark for current SDK version to capture any possible regressions with previous SDK versions
• MobileBERT model addition for MLPerf inference benchmark: ->
• Used SQUAD(Stanford Question Answer Dataset) as the source for groundtruth
• Developed F1 score verifier , using precision and recall, to compare the Tensorflow and SNPE generated output
• Throughput test framework development: ->
• Developed a framework in Python for parallel execution of multiple neural networks on different threads and different combination of processors , for measuring inference per second
• throughput, to ensure multiple neural networks are giving expected inference values when executed concurrently on mobile platforms

Sep 2019 – May 2021 · 1 yr 8 mos

• Worked in Modem MPSS team
• Internal contributions and good understanding of ETSI 3GPP and GSMA SGP specifications for Remote Sim Provisioning applications
• Added 5G support in Modem and contributed to multiple 5G developments in various devices in the transition from 4G to 5G.
• Contributed to various features on Embedded UICC technology, that involves usage of only 1 sim card having multiple operator specific profiles
• Worked as a Primary owner of Local Profile Assistant module, which does profile management in physical/virtual Embedded Sim cards.
• Managed Generic Bootstrapping Architecture module, that handles the authentication of sim card in secure applications like mobile banking, RCS.
• Contributed to design and working of SIM Subsidy and Remote SIM Unlock subsidy

Jul 2018 – Sep 2019 · 1 yr 2 mos

• Worked as a Secondary owner of Local Profile Assistant module, which does profile management in physical/virtual Embedded Sim cards.
• Worked in Generic Bootstrapping Architecture module, that handles the authentication of sim card in secure applications like mobile banking, RCS.
• Contributed to design and working of SIM Subsidy and Remote SIM Unlock subsidy

May 2017 – Jul 2017 · 2 mos · Hyderabad Area, India

• Developed a Machine Learning Application, currently used by the organization, that predicts whether any low, medium or high intent user will buy Microsoft Azure Subscription or not