Customer Support Tickets That Code

2026-07-01 · Source: The Information · Field: Technology & Digital — Artificial Intelligence & Machine Learning, Software Development & Engineering, Cybersecurity & Data Privacy · Depth: Intermediate, quick

Summary

Software developers are integrating large language models (LLMs) to automate the conversion of customer support tickets, detailing software issues or feature requests, directly into new code. This approach streamlines a previously lengthy process involving support teams, engineers, and manual coding. For instance, AI cybersecurity startup Semgrep routes customer bug complaints or feature requests to Anthropic's Claude model. Claude then generates fixes or new features within a sandboxed version of Semgrep's application. A human engineer subsequently tests these changes before they are merged into the main Semgrep app, which assists developers in scanning code for vulnerabilities. Currently, Semgrep applies this method to a handful of tickets monthly, with founder Isaac Evans anticipating rapid expansion.

Key takeaway

For AI Engineers or Directors of AI/ML evaluating ways to accelerate development cycles and improve customer response, consider implementing LLM-driven code generation from support tickets. You should establish a robust sandboxing and human review process to ensure code quality and security before deployment. This approach can significantly reduce the manual effort in addressing customer-reported issues and feature requests, scaling your team's responsiveness.

Key insights

LLMs are being used to automate the transformation of customer support tickets into functional code.

Principles

• Combine LLM strengths: coding and customer interaction.
• Isolate AI-generated code in a sandbox.
• Human oversight is critical for code quality.

Method

Customer complaints or feature requests are routed to an LLM, which generates code in a sandboxed environment. A human engineer then tests and merges the validated changes.

In practice

• Route support tickets to models like Claude.
• Implement sandboxed environments for AI code.
• Integrate human testing into the merge process.

Topics

• Customer Support Automation
• Code Generation
• Large Language Models
• AI in Software Development
• Anthropic Claude
• Semgrep

Best for: CTO, VP of Engineering/Data, AI Architect, AI Engineer, Software Engineer, Director of AI/ML

Related on AIssential

• The TechBeat: Can LLMs Generate Quality Code? A 40,000-Line Experiment (1/6/2026) — LLMs generate sloppy code over time, requiring formal analysis and multi-model reviews for quality assurance.
• LLM Vs AI: A Practical Guide to Differences, Use Cases, and Tools — Generative AI and LLMs are transforming enterprise operations through diverse applications, requiring careful selection and deployment.
• How Braintrust turns customer requests into code with Codex — High-speed AI code generation fundamentally transforms development workflows, enabling real-time customer iteration and autonomous problem-solving.
• Helping people write code again — LLM-assisted coding empowers experienced individuals to resume programming by reducing time barriers.
• Same Scrutiny, More Time: Eye Tracking Insights into Reviewing LLM-Labelled Code — Explicitly labeling code as LLM-generated increases review time without improving thoroughness, driven by perceived origin.
• Faster Code, Deeper Debt? A Multivocal Literature Review on Technical Debt and Its Early Signs in LLM-Assisted Software Development — LLMs accelerate traditional technical debt and introduce new, unique forms, necessitating novel management strategies and metrics.

Open in AIssential →

Editorial summary, takeaway, and curation by AIssential. Original article published by The Information.