I'm Praneeth — I build AI agents
and the infrastructure they run on.

By day I'm on the observability team at Aptean, keeping monitoring pipelines, centralized logging and alerting honest across a large fleet of enterprise products — OpenTelemetry, Prometheus, Grafana, Loki, Kubernetes, and the production troubleshooting that comes with them. Lately that includes pulling AI into the workflow: natural-language interfaces over operational data, AI-assisted troubleshooting, and an LLM-powered SQL agent.

The rest of the time I build. I run open-source models on GPU infrastructure (vLLM, LiteLLM, Ollama), maintain Magpie — an open-source control plane for OpenTelemetry collector fleets — and ship product experiments under InsightX Labs, which mostly means learning whatever the problem requires: backend architecture, Docker, reverse proxies, CI/CD, and the late-night production debugging that follows.

Most of my projects start with “can an agent do this?”. Some become products; some should have stayed weekend experiments. I talk about the interesting ones on VeepTalks, and when I'm away from a terminal I'm usually out with a camera looking for birds.