Sales ops · Production

From Slack Bot to Sales Brain: How Netguru Built the Omega AI Sales Agent

The problem

Netguru's sales team had critical information scattered across Slack, CRMs, call transcripts, and shared drives, slowing coordination and requiring repetitive manual effort for tasks like call prep, proposal review, and feature list generation.

Workflow diagram · grounded in source
1
User request via Slack
trigger
“Slack API – listens in channels, responds to mentions, and threads replies”
2
Routing Logic assigns agent
routing
“Routing Logic – Interprets the user's intent and assigns the task to the appropriate agent”
3
SalesAgent analyzes request
ai_action
“SalesAgent – Ensures each request aligns with the Sales Framework 2.0 by analyzing the user input and determining the appropriate next steps”
4
PrimaryAgent executes task
ai_action
“PrimaryAgent – Carries out the actual task based on the SalesAgent's analysis and the user's request, leveraging various tools (e.g., Google Drive) to generate a response”
5
CriticAgent validates output
validation
“CriticAgent – Reviews the PrimaryAgent's output and provides constructive feedback or validation to ensure quality”
6
Actionable response delivered
output
“delivering actionable insights, not just surface-level responses”
Reported outcome

Omega now supports the sales team at multiple stages of the sales process from within Slack, reducing repetitive work, adding process clarity, and improving access to information, with adoption described as frictionless.

Reported metrics
Expert call prep timesaving time and ensuring consistency
Repetitive sales workreduces repetitive work
Team adoption frictionadoption frictionless
Reported stack
AutoGenAWS LambdaAWS Step FunctionsTerraformAWS Systems Manager Parameter StoreSlack APIGoogle Drive APIApollo APIBlueDotLangfusePromptfooCircleCISlackApolloDrive
Source
https://www.netguru.com/blog/how-we-built-an-ai-agent
Read source ↗

Frequently asked questions

What did this team achieve with this AI workflow?

Omega now supports the sales team at multiple stages of the sales process from within Slack, reducing repetitive work, adding process clarity, and improving access to information, with adoption described as frictionless.

What tools did this team use?

AutoGen, AWS Lambda, AWS Step Functions, Terraform, AWS Systems Manager Parameter Store, Slack API, Google Drive API, Apollo API, BlueDot, Langfuse.

What results were reported?

Expert call prep time: saving time and ensuring consistency; Repetitive sales work: reduces repetitive work; Team adoption friction: adoption frictionless (source-reported, not independently verified).

How is this sales ops AI workflow structured?

User request via Slack → Routing Logic assigns agent → SalesAgent analyzes request → PrimaryAgent executes task → CriticAgent validates output → Actionable response delivered.

Related sales ops cases