Balaji NagarajanLeading AI, ML & data teams that ship at enterprise scale.https://balajin.dev/Why I started writinghttps://balajin.dev/blog/hello-world/https://balajin.dev/blog/hello-world/A home for my notes on AI, data, and building teams that ship.Tue, 07 Jul 2026 00:00:00 GMTNotes from GTC and Cloud Next: we have stopped talking about just modelshttps://balajin.dev/blog/two-phases-of-ai-gtc-2026/https://balajin.dev/blog/two-phases-of-ai-gtc-2026/Reflections from NVIDIA GTC and Google Cloud Next 2026 on the shift from AI models to AI systems that hold up in production.Wed, 15 Apr 2026 00:00:00 GMTMicro frontends: when the best thing to share is the app itselfhttps://balajin.dev/blog/micro-frontend-architecture/https://balajin.dev/blog/micro-frontend-architecture/Integration between teams keeps moving up the stack, from nightly files to streams to APIs. Micro frontends are the next step, and here is why they cut duplicate work and inconsistency.Mon, 22 Jul 2024 00:00:00 GMTScaling an ML algorithm with Spark as distributed computehttps://balajin.dev/blog/scaling-ml-solution-with-spark/https://balajin.dev/blog/scaling-ml-solution-with-spark/An ML algorithm written in Python had to run over tens of millions of items. Here is why I ran it as a distributed Spark job instead of an API or an event stream.Thu, 18 May 2023 00:00:00 GMTPerformance testing, part 2: publishing k6 metrics with InfluxDB and Grafanahttps://balajin.dev/blog/k6-metrics-influxdb-grafana/https://balajin.dev/blog/k6-metrics-influxdb-grafana/Once k6 was running, results vanished after each run. Here is how InfluxDB and Grafana turned load testing into something we could track and share.Sat, 20 Nov 2021 00:00:00 GMTWhy k6 for performance testing over JMeter and LoadRunnerhttps://balajin.dev/blog/why-k6-over-jmeter-loadrunner/https://balajin.dev/blog/why-k6-over-jmeter-loadrunner/Moving load testing into code with k6 so any engineer could run it, and why it beat the traditional tools for our workload.Mon, 20 Sep 2021 00:00:00 GMTWhy Druid: building a real-time analytics product for tens of thousands of usershttps://balajin.dev/blog/why-druid-real-time-analytics/https://balajin.dev/blog/why-druid-real-time-analytics/How I evaluated MongoDB, Cassandra, HBase, and Apache Druid for a sub-second, high-concurrency analytics product, and why Druid won.Tue, 15 Jun 2021 00:00:00 GMTWhy I chose Micronaut over Spring Boot (and Kotlin along the way)https://balajin.dev/blog/why-micronaut-over-springboot/https://balajin.dev/blog/why-micronaut-over-springboot/Picking a JVM framework for a high-throughput, low-latency data product, and why Micronaut and Kotlin were the right call.Wed, 10 Mar 2021 00:00:00 GMTReporting Kafka and Spark streaming metrics with a custom Spark listenerhttps://balajin.dev/blog/spark-streaming-metrics-listener/https://balajin.dev/blog/spark-streaming-metrics-listener/There is no ready-made dashboard for Kafka to Spark streaming health, so I built a listener to publish the metrics to InfluxDB and Grafana.Sun, 14 Apr 2019 00:00:00 GMT