Four members of the bbv AI Hub development team Thomas (consultant), Michelle (engineer), Noah (junior engineer), and I (architect) each presented how AI has changed their role. Marius Högger moderated.
Thomas dictates observations in Swiss German into the AI Hub's speech-to-text function immediately after customer workshops, capturing unstructured thoughts while the room's posters and post-its are still visible as memory aids. An AI assistant transforms that transcript into structured epics and requirements with acceptance criteria. Photos of post-it walls are fed directly to the model, which clusters and summarizes them. The same pipeline generates Mermaid diagrams and unit test skeletons from the resulting requirements. The models also perform gap analysis: given a set of requirements, they propose missing non-functional requirements and flag contradictions.
Michelle identified context management as the central challenge in AI-assisted coding. Context overload feeding the entire codebase or all open files produces vague, unspecific output. Context starvation missing project guidelines or language conventions produces generic code that doesn't fit. She manually selects exactly which code snippets to include, opens a new chat when switching topics, and disables web search by default (enabling it only for specific needs like tracking down library bugs). For recurring projects, she creates dedicated AI assistants with pre-loaded coding guidelines, saving the repetitive setup of each conversation.
Noah joined the team at the start of the year and used AI tools extensively to navigate an unfamiliar codebase. He connects the codebase to the AI Hub and queries it for abstract concepts "how do we handle data persistence?" letting the model search across all files and return both an explanation and the relevant source locations. For learning new concepts outside the codebase, he prompts the agent with a knowledge-tree metaphor: start with foundational concepts (trunk), proceed to specifics (branches and leaves), and do not advance until he confirms understanding. He also uses a GitHub Actions integration that runs an LLM review on pull requests before human review, catching security issues and giving an early quality signal.
My workflow operates across three abstraction levels, each requiring different tools. Deep Research (Gemini, GPT) surveys the landscape of available technologies for a given problem. I constrain it to trusted sources arXiv for papers, specific subreddits for practitioner experience, GitHub for source code to avoid low-quality web content contaminating the results. One level down, an IDE-integrated agent like Junie dives into specific open-source codebases to verify whether a library actually supports what its documentation claims. This runs asynchronously: I issue a question, continue other work, and review the findings minutes later. At the most specific level, I use a large-context model (Gemini with 1M+ tokens) and feed it the entire AI Hub codebase condensed via Repomix, which strips implementation details but preserves function signatures. With the full project in context, the model identifies exactly where a new component should be integrated. For actual code generation, I switch to Claude Sonnet with a narrow, manually curated context the large-context overview is too broad for precise code output.
After a customer workshop, speech-to-text tools capture unstructured observations in dialect. AI assistants then transform these transcripts, along with photos of posters and post-its, into structured epics, requirements with acceptance criteria, and even unit test skeletons. The models also identify gaps, contradictions, and missing non-functional requirements.
Context management. Too much context (e.g. feeding an entire codebase) makes responses unspecific. Too little context (missing project guidelines or language conventions) produces generic code. The solution is manual context selection: choosing exactly which code snippets to include, starting new conversations when switching topics, and creating project-specific AI assistants with pre-loaded guidelines.
AI tools can search and explain a large codebase on demand for example, answering how data persistence works across the project. For learning new concepts, prompting the model with a learning style (e.g. a knowledge tree from fundamentals to specifics) and instructing it to pause between concepts until confirmed keeps the pace manageable. AI-powered PR reviews also provide early feedback before senior review.
Deep Research tools (Gemini, GPT) provide the big-picture overview of available technologies for a given problem, constrained to trusted sources like arXiv, specific subreddits, and GitHub repositories. IDE-integrated agents like Junie then dive into specific open-source codebases to verify capabilities. Finally, a large-context model (1M+ tokens) receives the entire project codebase to identify exactly where a new component should be integrated.
For coding: Claude Sonnet 3.7 with reasoning enabled. For research: Google Gemini Deep Research or Perplexity. For general knowledge questions: any standard model without reasoning, since reasoning models like o3 tend to hallucinate more on factual queries. Avoid the free tier of ChatGPT for coding as it defaults to GPT-4 Mini.
.
Copyright 2026 - Joel P. Barmettler