Keshav Kumar

Head of Serverless Kafka

Confluent

May 2026 - Current (4 months)

We make Confluent 10x more elastic than Apache Kafka. Learn More: https://www.confluent.io/blog/cloud-native-kafka-kora-vldb-award/

Engineering Manager

Confluent

May 2022 - Current (4 years 4 months)

Building the Control Plane for Confluent Platform. On-Prem Observability & Management using Confluent Control Center. https://docs.confluent.io/platform/current/control-center/index.html Cloud Observability for On-Prem clusters using Confluent Health+ https://docs.confluent.io/platform/current/health-plus/index.html

Engineering Manager, Kafka Control Plane & Observability

Confluent

May 2022 - May 2026 (4 years 1 month)

Building the Control Plane for Confluent Platform. On-Prem Observability & Management using Confluent Control Center. https://docs.confluent.io/platform/current/control-center/index.html Cloud Observability for On-Prem clusters using Confluent Health+ https://docs.confluent.io/platform/current/health-plus/index.html

Tech Lead Manager - ML Ops Engineering

Domino Data Lab

Jan 2021 - Mar 2022 (1 year 3 months)

Domino's enterprise data science platform is used by 20% of the Fortune 100 companies. It competes with giants like AWS Sagemaker and Azure ML. I am part of the Model Ops engineering which is responsible for publishing, hosting and monitoring of machine learning models. Model Hosting: I evaluated adoption of kubernetes native third party hosting frameworks like seldon, kubeflow/kserve for hosting models. Model Monitoring: Re-engineered the storage layer of model monitoring platform to solve problems of query performance, scale and availability by adopting, tuning and operationalizing TimescaleDb, a time series database on top of Postgres.

Tech Lead - ML Ops Engineering

Domino Data Lab

Jan 2021 - Mar 2022 (1 year 3 months)

Domino's enterprise data science platform is used by 20% of the Fortune 100 companies. It competes with giants like AWS Sagemaker and Azure ML. I am part of the Model Ops engineering which is responsible for publishing, hosting and monitoring of machine learning models. Model Hosting: I evaluated adoption of kubernetes native third party hosting frameworks like seldon, kubeflow/kserve for hosting models. Model Monitoring: Re-engineered the storage layer of model monitoring platform to solve problems of query performance, scale and availability by adopting, tuning and operationalizing TimescaleDb, a time series database on top of Postgres.

Engineering Manager, Growth

Vahan Inc

Mar 2020 - Sep 2020 (7 months)

Vahan's jobseeker platform is enabling the next billion internet users in emerging markets connect to relevant opportunity. Bootstrapped the engineering processes to support two fold growth of the engineering team. Worked with consultants in data engineering, product management and UI/UX to fast-track product development.

Engineering Lead

Vahanai

Mar 2020 - Sep 2020 (7 months)

Vahan's jobseeker platform is enabling the next billion internet users in emerging markets connect to relevant opportunity. Bootstrapped the engineering processes to support two fold growth of the engineering team. Worked with consultants in data engineering, product management and UI/UX to fast-track product development.

Software Engineer

Nov 2017 - Nov 2019 (2 years 1 month)

Part of Communications Platform Team at LinkedIn which is responsible for sending communication to around 600 million LinkedIn members over multiple channels like email, push, sms with support for both synchronous and asynchronous delivery. It serves requests at around 600k QPS across 4 data centres across the world. It provides an ecosystem to other teams within LinkedIn to self service their use cases like campaign creation and management, tracking conversions, monitoring failures etc.

Software Engineer, Notification Platform

Nov 2017 - Nov 2019 (2 years 1 month)

Part of Communications Platform Team at LinkedIn which is responsible for sending communication to around 600 million LinkedIn members over multiple channels like email, push, sms with support for both synchronous and asynchronous delivery. It serves requests at around 600k QPS across 4 data centres across the world. It provides an ecosystem to other teams within LinkedIn to self service their use cases like campaign creation and management, tracking conversions, monitoring failures etc.

Software Engineer, Apps Platform

Directi

Aug 2015 - Nov 2017 (2 years 4 months)

Flock is a SaaS-based enterprise communication application aimed at disrupting the communication space with its rich and powerful integrations. 1. Design and development of the web service that powers the flock app store where users can configure integration with the third-party applications like github, trello, asana and so on. This allows users to get notified of events and access information from within Flock. 2. Leveraging the Flock developer platform to build deep integration with third-party applications.(Eg: uber, google drive, google calendar etc)

Software Engineer

Directi

Aug 2015 - Nov 2017 (2 years 4 months)

Flock is a SaaS-based enterprise communication application aimed at disrupting the communication space with its rich and powerful integrations. 1. Design and development of the web service that powers the flock app store where users can configure integration with the third-party applications like github, trello, asana and so on. This allows users to get notified of events and access information from within Flock. 2. Leveraging the Flock developer platform to build deep integration with third-party applications.(Eg: uber, google drive, google calendar etc)

Quantitative Analyst Developer

Goldman Sachs

Jun 2014 - Aug 2015 (1 year 3 months)

Contributed to following divisions: 1. Quantitative Investment Strategies - Data Platform The data platform gets time sensitive market data from various sources and curates them into its data stores which act as a single source of truth for various systems used by the Quantitative Systems to analyze and execute trades. 2. Investment Guidelines Platform: Guidelines Platform facilitates real-time, distributed, high volume, low latency rule evaluation service to the traders. I worked closely with the Strategists to improve the rule evaluation algorithm to increase accuracy.

O

Member Technical Staff, Oracle Cloud Infra

Oracle Server Technologies

Jun 2012 - May 2014 (2 years)

Contributed to framework development for Oracle Enterprise Manager Cloud Control. Oracle Enterprise Manager (OEM or EM) is a set of web-based tools aimed at managing and monitoring services running on over 40k VM's on the Oracle Cloud. Oracle Enterprise Manager performs much of its activity through intelligent agents refers to as Oracle Management Agents. These run as autonomous proxy processes on a managed node, and perform execution and monitoring tasks for Oracle Enterprise Manager, communicating over http.

Member Technical Staff

Oracle

Jun 2012 - May 2014 (2 years)

Contributed to framework development for Oracle Enterprise Manager Cloud Control. Oracle Enterprise Manager (OEM or EM) is a set of web-based tools aimed at managing and monitoring services running on over 40k VM's on the Oracle Cloud. Oracle Enterprise Manager performs much of its activity through intelligent agents refers to as Oracle Management Agents. These run as autonomous proxy processes on a managed node, and perform execution and monitoring tasks for Oracle Enterprise Manager, communicating over http.

About

Detail

Timeline

Résumé