Satish Raghunath Email & Phone Number

San Francisco, California, Us

Heading Hyperforce infrastructure engineering services that power our CRM products.We build cutting edge software that helps us deliver Software-as-a-Service with high availability, satisfying complex compliance requirements for a global customer base. The Hyperforce platform is built on top of public cloud primitives to meet our customers' needs.My team drives network security and software services that are at the heart of Hyperforce. These products provide content delivery (CDN), private connectivity (VPN), policy based filtering, access control and protections against L3/L4 and L7 attacks (WAF). We build software that realizes the best of breed functionality with an appropriate mix of off-the-shelf open source / commercial products and in-house innovation.Our software services are mission critical to our business and to our customers. To do this at scale and deliver innovation at a rapid pace, we invest heavily in autonomous systems and safe change practices.I also help our front line teams with customer conversations providing insight into complex technical problems as well as helping provide our customers visibility into issues and problems that sometimes affect our systems.

San Francisco, California, Us

Driving software engineering for the Core App infrastructure. We are building Edge services and load balancing software that helps deliver superior experience to a global user base. Salesforce Edge is optimized to help improve delivery of dynamic content. The Edge service is like a Content Delivery Network (CDN) with a few differences - it targets a subset of performance enhancing features that is suitable for compliance sensitive data. To meet our scaling and compliance needs, we have an optimized control plane that caters to customizations across millions of domains.Taking forward the innovations in Twin Prime, we generate data-driven insights from network performance metrics to improve the traffic path.With software load balancing, we are helping adoption of the latest in the open source community. We rely on Envoy, Istio and Nginx to help achieve application objectives such as hit-less continuous deployment.

San Francisco, California, Us

Infrastructure Engineering focusing on network edge, cloud and mobile app performance. Deploying machine learning to help drive better app performance.Helped roll out the Twin Prime mobile optimization software to all Salesforce iOS/Android customers. Delivered download time reductions in the range of 15-25% worldwide for marquee brands.Helped drive a custom built network edge service - the Salesforce Edge - that provides TLS termination and Caching closer to our end-users. Purpose built to exceed Salesforce security standards and suitable for compliance and privacy sensitive traffic. Optimized for dynamic content. Deployed worldwide and delivering 2+ seconds in latency cuts to some of the largest enterprise customers of Salesforce Sales and Service Cloud products.Driving a data driven approach to improving infrastructure software - introducing metrics to understand infrastructure better, building visibility and generating insights from metrics. Hiring and retaining a strong developer team that brings deep knowledge of the latest in infrastructure technologies.Contributed to the internal DNS architecture and strategy, public cloud network strategy.

United States, Us

Redwood City, California, Us

Twin Prime was acquired by Salesforce!Twin Prime built a service to make mobile apps run faster. Performance on mobile is a complex problem due to the large number of factors that are relevant - the huge diversity in mobile devices, network types, wireless channel variability, breadth of apps and even the backend. We used machine learning to not only help customers answer why their mobile apps are slow, but also to automatically implement software strategies to mitigate performance issues. We deployed this solution via a mobile SDK that runs on Android and iOS in conjunction with a cloud-based component called the Accelerator. The machine learning framework chose the parameters to achieve best network performance: request concurrency, pipelining and failure detection on the client side; congestion control, loss recovery and transport protocol strategies on the server side. Our technology was purely at IP and TCP layers and hence worked with all kinds of content - even encrypted content without requiring TLS termination.At the time of acquisition, several large customers in news publishing, mobile commerce and travel relied on our technology to delight their end-users on mobile. The technology delivered dramatic improvements on a global scale across millions of mobile users.At Twin Prime, I not only played a key role in fundraising, conceiving and implementing the technology, I drove the technology direction and strategy, hired / mentored / managed the top-notch engineering team, and led many of the key customer engagements. We raised multiple rounds of financing from top-tier venture firms. We built out a globally distributed cloud presence, a large scale data pipeline that harvests metrics from millions of mobile handsets, a machine learning framework continually took in data and deployed custom policies out into the global network, a visualization dashboard and finally a multi-cloud configuration management system.

Sunnyvale, Ca, Us

IP Layer 3-7 services for edge and core routers. Design and implementation of a high performance platform for rich service delivery in routers. Enabled multiple third-party and Junos-based applications to co-exist within the same service complex. Helped define Juniper-wide service framework API as part of the effort to build a services platform across the company. Enabled TCP proxy API for various services delivered on MX platform - Ad optimization, HTTP payload rewrite, application gateways for NAT.

Sunnyvale, Ca, Us

Design and implementation of the Jflow-v9 service for the M-/MX-/T-series routers. Delivered packet path functions for Juniper's session border controller voice-over-IP product - fine-grained large-scale policers, NAT, control plane integration for call setup. Design of IPv6 transition technologies - Carrier-grade NAT, application gateways (ALGs), Protocol translation strategies (IETF BEHAVE) like DS-Lite.

Ca

Research and prototype, solution for dynamic resource allocation (layer 1-7, Bandwidth, Protection etc) for networks in the presence of heterogeniety in network technologies (Ethernet, SONET/SDH, long-haul optical, Wireless etc) and administrative authority (inter-domain issues). Cross-layer resource orchestration to make networks application-aware.

Dallas, Tx, Us

IP MPLS VPN design and architecture; large-scale measurement study of VPN traffic for structural characterization; strategies for dynamic adaptation of bandiwdth allocations while reconciling SLA requirements.

Us

Design and build a tool to recreate traffic from packet traces. Spoofing layer-specific information, accelerating or decelerating actual trace durations, support for popular trace formats.

Provo, Ut, Us

Porting FDDI component of RMON agent from NetWare to Windows NT. Automated test scripts for RMON agent. Design for RMON-II support. Support for network-based license verification. Java network service monitor.

Ms, Ph.D., Computer Networking

B.E., Computer Engineering

Education record

Certificate In Business Administration, Business Administration