RESEARCHER | ENGINEER | PLATFORM DESIGNERHello, I am currently searching for new opportunities!I am interested in designing and building customer focused products, while taking up challenging problems. My experience revolves around designing and architecting innovative solutions for Cloud, Edge, Geo-Distributed systems and AI, ML, Big Data. My work consistently bridges these two domains, and I remain dedicated to staying current with advancements in both areas. I hold a PhD (in one months) from the renowned Inria Labs and IMT Atlantique, France. I have experience in managing a 2-10 amount of people at various roles through my carrier. My experience has helped me to find solutions for challenging problems. My Expertise:Cloud Services:✓ AWS✓ Azure✓ GCP✓ StackMLOPS:✓ Kubeflow, Airflow, MLflow✓ ML training environment optimization✓ ML model Deployment✓ Serving models✓ Tensorflow, Pytorch, Seldon, Mxnet, OnnxDevOps:✓ Kubernetes✓ Docker✓ Ansible✓ jenkins✓ Git (GitLab & GitHub)✓ Nginx, HAproxy✓ LinuxBig Data:✓ Hadoop✓ Spark✓ Kafka✓ HiveI am looking forward to working in these domains, if you are seeking for a person with positive attitude and leadership quality, ping me here.
-
Freelance Software DeveloperNantes, Fr
-
Freelance Software DeveloperNantes, Fr
-
Research And Development EngineerImt Atlantique Apr 2024 - Aug 2024Nantes, Pays De La Loire, France• Developed a framework to enable geo-distribution of applications (Kubernetes and Openstack) without modifying internal code, increasing scalability and resilience.• Designed and developed a P2P architecture for geo-distributing Kubernetes and Openstack, it used Rabbit MQ and Couch DB.• The design followed a Service Oriented Architecture and used modular components connected with REST API for scalability.• The approach was tested under scenarios with large number of transactions and across different geo-graphically distributed areas.• Automated infrastructure provisioning with IAAS tools, Ansible and Terraform, reducing setup time and enhancing scalability.• Led a team of 2 engineers to create the framework, with development in Golang programming.- Tools: Kubernetes, go-lang, couchDB, Ansible, Terraform, Git, Docker, Cloud, Edge -
Doctoral ResearcherInria Apr 2021 - Mar 2024Greater Nantes Metropolitan Area• Researched and developed a new approach to enable geodistribution of applications (e-commerce, text editors, cloud tools) without modifying internal code, increasing scalability and adaption.• The platform is generic to any application and can be reused.• Proposed a new sharding based collaboration model, solving issues of replication, to geo-distribute any application or resources.• Proposed a new consistency approach apart from the popular Consensus (RAFT, Paxos or Leader based) and CRDT, to facilitate synchronization external to the geo-distributed resources.• Proposed a dependency management approach to ensure geodistribution between Cloud, Edge (CDN) are seamless.• Combined two programming models: Object-oriented and Aspect, to provide multiple flexible and customized geo-distribution strategies.• Supervised the project, providing the technical stack design and leadership, conducting code reviews. -
Mlops Platform EngineerAivid Techvision 2020 - 2021India• Developed a scalable MLOps platform tailored for computer vision applications, leveraging OpenCV, OpenVINO, PyTorch, and TensorFlow for diverse model support.• Established real-time video streaming data pipelines using Kafka, creating seamless data ingestion and model inferencing across cloud and edge environments (bare metal servers).• Storage was facilitated for these data across local object stores with MinIO and MongoDB, Redis was used for indexing.• Deployed model-serving APIs on Kubernetes, with integration options like TF Serving, OpenVINO, and Flask, maximizing platform flexibility for client needs.• Automated CI/CD with Ansible and GitHub Actions, ensuring continuous deployment and comprehensive monitoring using Prometheus and Grafana.• Utilized service meshes like Istio and Consul to collaborate components across multiple locations and to coordinate internal model training with multiple containers. -
Mlops Platform EngineerReliance Industries Limited 2019 - 2020Mumbai, Maharashtra, India• Developed an MLOps platform for the data science team, integrating data ingestion and model orchestration using Kubeflow and Airflow.• Designed a custom engine for model training and hyperparameter tuning, leveraging Katib, Kubernetes and OpenFaaS for efficient parallel processing.• Created a highly available and robust serving system with Kubernetes and models exposed with PyTorch and TF serving.• Integrated multiple stages of the pipeline using tools like gitlab, Kubeflow, Kubernetes, TF serving, Pytorch serving, kflow, Kafka, Hadoop and Spark.• Implemented monitoring solutions with the ELK stack (Elasticsearch, Logstash, Kibana) on Kubernetes, improving model visibility and error tracking.• The infrastructure consisted of a collaboration between Bare metal servers, Azure and Google cloud.• Received the “Employee of the Year” award for designing a highperformance MLOps pipeline, reducing latency and enhancing reliability in model deployment.
Geo A. Skills
Geo A. Education Details
-
Computer Systems Networking And Telecommunications
-
Big Data & Ai -
Computer Engineering
Frequently Asked Questions about Geo A.
What is Geo A.'s role at the current company?
Geo A.'s current role is Freelance Software Developer.
What schools did Geo A. attend?
Geo A. attended Imt Atlantique, Vellore Institute Of Technology, Christ University, Bangalore.
What skills is Geo A. known for?
Geo A. has skills like Amazon Web Services, Data Science, Microsoft Azure, Kubeflow, Cloud Native Architecture, Devops, Hadoop, Artificial Intelligence, Google Cloud Platform, Programming, Machine Learning Algorithms, Docker.
Not the Geo A. you were looking for?
Free Chrome Extension
Find emails, phones & company data instantly
Aero Online
Your AI prospecting assistant
Select data to include:
0 records × $0.02 per record
Download 750 million emails and 100 million phone numbers
Access emails and phone numbers of over 750 million business users. Instantly download verified profiles using 20+ filters, including location, job title, company, function, and industry.
Start your free trial