DrDroid at DevOps 2.0 Confex & Awards 2026

DrDroid joined the 5th DevOps 2.0 Confex & Awards as Gold Partner, with a talk on who stays accountable once autonomous agents start doing production work.

DrDroid was Gold Partner at the 5th edition of DevOps 2.0 Confex & Awards, one of India's larger gatherings of DevOps, SRE and platform engineering teams, held at Taj MG Road, Bengaluru under the theme "The Next Wave of DevOps: Agile, Efficient, Impactful".

We were there to talk about a shift every on-call team is already feeling: AI can now investigate production issues end to end, and the interesting question is no longer whether it can, but what that does to ownership.

Autonomous agents are coming. Who is accountable?

Our talk argued that accountability does not disappear when an agent does the investigating — it moves up a level. Engineers stop being accountable for running the queries and become accountable for the agents running them.

In practice that means three things: evaluate the capability of the agents your team is using, plug the higher-order KPI back to humans, and feed what you learn from real incidents into improving the agents. Skip any one of those and you have automation nobody owns.

What we demoed at the booth

Between sessions we ran live investigations against real alerts. No runbooks, no prepared data — engineers described a failure they had debugged the week before and we pointed DrDroid at the same class of problem.

• Trace analysis and log correlation across services to find a latency regression
• Root cause for a memory spike behind a CrashLoopBackOff, without a runbook
• A cascading failure traced back through the service that actually caused it
• Alert grouping collapsing an alert storm into a single issue with one owner

What we heard from engineering teams

The same theme came up in nearly every conversation: teams are far more comfortable automating investigation than automating action, and they want the handoff between the two to be explicit rather than implied.

The second theme was context. Nobody doubts the data exists — the problem is knowing which of forty dashboards matters for this alert, at the hour when nobody is thinking clearly.

What DrDroid does

DrDroid is an AI SRE agent for infrastructure teams. It builds a knowledge graph of your stack from your existing telemetry, investigates alerts as they fire, and hands your on-call engineer a root cause instead of a dashboard to go read.

AI investigations

Alerts are investigated as they arrive, with the reasoning shown rather than a bare answer.

Alert grouping

One root cause becomes one issue with one owner, instead of dozens of near-identical pages.

Self-learning agent

Every investigation feeds back in, so the agent gets sharper on your systems specifically.

80+ integrations

Grafana, Kubernetes, AWS, Sentry, Datadog and the rest of the stack — all read-only.

Runs where your data lives

Self-hosted in your VPC or on-prem, SOC 2 Type II, with air-gapped deployment supported.

Missed us in Bengaluru?

The talk is up on our YouTube channel, and the team is happy to run the same live investigation against your stack instead of ours.