IT Specialist II / Automation Builder / Internal platform and workflow engineering

Root Insurance Automation

Built internal automation that turned Slack issue intake into structured ServiceNow ticket creation using AWS Lambda and a Python Flask service, with attachment support, decision-tree routing, and smarter handling for account lockout cases.

EngineAWS Lambda + Python Flask integration

TimelineFebruary 2024 - March 2026

Pathway to App Store Connect

Engineered an AWS Lambda and Python Flask integration that created ServiceNow tickets directly from Slack intake.
Added attachment upload support so the ticket payload preserved the context users were sending with their requests.
Built decision-tree logic to classify tickets based on issue type, with specific filtering for different kinds of lockouts.
Piloted a retrieval-augmented API workflow and refined the prompt layer so claims-training users received company-specific guidance instead of generic model output.

Tools and context

Tech: AWS Lambda, Python, Flask, Slack API, ServiceNow API, Prompt engineering, Retrieval augmentation
Constraint: The workflow had to reduce manual triage without losing ticket quality, attachment fidelity, or the routing detail IT needed to respond correctly.
Outcome: Reduced friction between user requests and IT operations while creating more structured tickets and more reliable training outputs in an enterprise environment, alongside broader AWS workplace experience relevant to government and enterprise solutions.

Breakdown

Decision-making and implementation

Problem

IT intake was too manual and too inconsistent

Users were already communicating in Slack, but turning those messages into useful ServiceNow tickets still required manual handling, interpretation, and follow-up.

Decision

Automate structure, not just submission

I treated the problem as a translation layer between user behavior and service-management requirements, which meant handling attachments, routing logic, and lockout distinctions rather than just pushing raw text across systems.

Result

The system became more useful to both users and IT

The result was a cleaner, more actionable intake path for IT and a better user experience for internal teams, alongside more targeted retrieval and prompt behavior for training workflows.

Back Home