Exla GTM brain

category

market profile

• text: Engineering teams need to connect production failures to affected customers and understand impact, rather than only collecting error logs.
• text: AI SRE and observability vendors compete to speed incident investigation with evidence-backed diagnosis and remediation.

personas

• title: Startup CTO or engineering founder
  pains:

  • Customers discover broken workflows before the team notices.
  • A small team cannot manually interpret every log line and track every affected account.
• title: On-call software engineer
  pains:

  • Must sift through noisy production logs to find the failure and its likely cause.
  • Needs enough context to reproduce and repair a customer-specific issue quickly.
• title: Engineering lead responsible for customer reliability
  pains:

  • An exception count alone does not reveal which customer or how many users are affected.
  • Prioritizing a fix requires both customer impact and evidence that engineers can inspect.

icp

• text: B2B software companies running production services with identifiable customer accounts and logs; no employee count or revenue threshold is specified.
• text: Engineering teams using coding agents such as Claude Code or Cursor and willing to connect log context through an explicitly approved MCP setup.

message pillars

• text: See what broke, which customer was affected and why in a plain-language summary.
• text: Ground the diagnosis and proposed fix in underlying production log lines.
• text: Set up with an existing stack or a coding agent, with user approval for sensitive connections.
• text: Detect customer-impacting issues before customers report them.

objections

• objection: We already use Sentry or a log platform.
  response: Show how Libra adds customer impact and a diagnosis backed by log lines, and verify integration with the existing stack.
• objection: Can an AI agent access source code or modify my coding-tool configuration without consent?
  response: The setup flow explicitly requests separate approval for MCP configuration and for sending source files to a model.
• objection: Will an AI diagnosis be reliable enough to act on?
  response: Inspect the cited log lines and review proposed changes before deploying; the output is a starting point for investigation.

competitors

• name: Sentry
  why: Error monitoring incumbent explicitly compared in Libra's positioning.
• name: Resolve AI
  website: https://media.website-prod.resolve.ai/2025_SRE_Buyers_Guide_Single_Pg_v3_112625_58319e8071.pdf
  why: AI SRE investigation and incident diagnosis alternative.
• name: Cleric
  why: AI SRE and observability agent for incident investigation.

keywords

• AI customer monitoring
• customer impact production logs
• AI log analysis
• AI SRE agent
• incident root cause analysis
• customer specific error monitoring
• coding agent production logs

voice

• Direct and developer-focused
• Plain-spoken and concrete
• Evidence-led

signals

• text: A B2B SaaS team has customer-specific integration failures that are buried in production logs.
• text: The team is adopting coding agents and wants production evidence accessible in its development workflow.
• text: Customer reports are arriving before engineers can identify affected accounts and prioritize a fix.

stack

• Claude Code
• Cursor
• Codex
• MCP

pricing

motion

channels

• SEO
• self-serve website

confidence

Founders

• name: Viraat Das
  title: Founder/CEO
  role: CEO
  bio: CEO @ Exla. Previously machine learning engineer @ Amazon.
• name: Pranav Nair
  title: Founder/CTO
  role: CTO
  bio: CTO at Exla. Previously an OS engineer at Apple leading sleep/hibernation for all Apple devices.