Frame tasks as a chain of "and then" actions an infinitely staffed team would perform. For example, a customer query in Slack is answered, "and then" AI turns it into a help article, "and then" it becomes SEO content. AI makes these previously cost-prohibitive workflows achievable.
To overcome engineering hesitancy about giving customer-facing teams codebase access, highlight the benefit to engineers themselves. Position it as a solution that eliminates their need to answer constant, last-minute pings and DMs, freeing them from being an information bottleneck.
Instead of manually performing tedious tasks like 'git pull' across 15 repositories, use an AI assistant like Claude Code to instantly write a script. This automates environment setup and maintenance, ensuring local code is always up-to-date with minimal effort.
Instead of meticulously organizing information, teams can let AI query across code repositories, Confluence, and Slack. This allows for more operational chaos, as AI can find and synthesize information regardless of where it's stored, reducing the administrative burden of knowledge management.
Field engineers can bypass documentation limitations by querying the entire codebase with AI tools like Claude Code. This provides detailed, step-by-step answers that public docs lack, directly addressing complex customer problems and reducing reliance on the engineering team.
Maintain a simple Confluence page with bulleted lists of each enterprise customer's unique technical needs (e.g., security, secrets management). This page serves as powerful context for an AI assistant, enabling it to generate highly tailored and trustworthy deployment guides for complex environments.
While many teams use AI to accelerate product development, a key advantage lies in using it to improve customer interactions. Providing customized deployment plans and deep technical answers shows customers you understand their specific needs, building trust and positioning your team as a superior partner.
LLMs' effectiveness in understanding code means even non-technical roles must develop basic coding literacy. Being comfortable opening an IDE and understanding Git basics is becoming a fundamental hard skill, as code becomes the primary medium for communicating with AI assistants.
When working with multiple repositories, opening the entire project directory in your IDE allows AI tools to traverse all repos. This provides more contextualized answers to complex questions that span multiple services, avoiding siloed analysis and improving AI assistant performance.
When an AI's response is questionable, go beyond simple re-prompting. Use meta-prompts that explicitly instruct the model to increase its reasoning effort, such as "Think hard about why this is right" or asking for its sources. This can uncover new insights and improve output quality.