30 January 2025 · 8 min read
AI Upskilling by Function: What Each Team Should Actually Learn
A one-size-fits-all AI training program trains no one well. Here's a practical breakdown of what Engineering, Product, GTM, Growth, and Data teams need to learn - and what they can skip.
By The AI Internship Team, TAI Labs
From the archive. Originally published 2025-01-30. Tool details, examples and offers reflect that publication date. See our current guides for newer material.
Why function-specific AI training outperforms generic programs
The most common failure mode of enterprise AI training is abstraction: everyone sits through the same LLM overview, the same prompt engineering basics, the same "AI landscape" slides - and leaves without a single workflow they can apply on Monday morning.
Effective AI upskilling is anchored to the specific tasks each function performs. Here's what that looks like by team.
Engineering teams
Core tools: Claude Code, Cursor, GitHub Copilot, LangChain/LlamaIndex, Ragas
Key skills: Agentic coding workflows, RAG architecture, evals design, LLM infra and cost control, multi-agent orchestration
Skip: Generic "introduction to LLMs" content. Engineers need to work with models at the API level from day one.
Product teams
Core tools: Claude (Projects), Lovable, v0, Cursor, Linear AI
Key skills: AI spec writing, prototype-to-pilot handoff, evals for product quality, prompt pipeline design, AI feature scoping
Skip: Deep technical model internals. PMs need to understand capabilities and limitations, not architecture.
GTM & Sales teams
Core tools: Clay, Apollo AI, Instantly, Claude, Gong AI
Key skills: AI-enriched prospecting, personalised outbound at scale, discovery call intelligence, pipeline forecasting with AI assist
Skip: Coding and model fine-tuning. GTM upskilling should be entirely workflow and tool-based.
Growth & Marketing teams
Core tools: Claude, Perplexity, Midjourney/Firefly, n8n, Make
Key skills: Programmatic content at scale, AI-assisted SEO, lifecycle automation, campaign intelligence
Skip: Deep engineering content. Growth teams need to understand what's possible and how to direct engineers, not build the systems themselves.
Data teams
Core tools: Claude (for SQL and Python), Cursor, dbt, LLM eval frameworks
Key skills: LLM evals pipelines, production ML monitoring, AI-assisted data modelling, LLM cost and quality analytics
Skip: Generic "intro to AI" material. Data teams need the production ML ops and evals content.
Custom-designed tracks for every function
We design upskilling programs per function and run them as aligned cohorts. Book a discovery call →