AI Assistance for Software Teams: The State of Play • Birgitta Böckeler • GOTO 2024

AI Assistance for Software Teams: The State of Play • Birgitta Böckeler • GOTO 2024

by Birgitta Böckeler

Applied Sciences & Engineering Computing & Cybersecurity UMComputer programmingUMZSoftware Engineering
🎙 Birgitta Böckeler 👥 1.1M 📅 September 10, 2025 ⏱ 45 min 👁 2K 📄 expert opinion 🧭 2026-08-03
Available in: English (current) Français

Keywords

coding assistantsAI toolingsoftware deliverycontext orchestrationmodel choices

Summary

Birgitta Böckeler, Technical Principal at Thoughtworks, presents an overview of AI assistance for software teams, focusing on coding assistants. She categorizes AI tooling into archetypes: coding assistants, extensions to delivery tool chains, and other tools for tasks like requirements analysis. She notes that coding assistants are the most mature and widely adopted, while other categories are less developed. The talk covers core features of coding assistants (inline suggestions, integrated chat), model choices (including hosting options and data privacy), and advanced features like context orchestration (e.g., indexing codebases). She discusses the importance of model quality, citing GPT-4 and Claude Sonnet as top performers. She also touches on future frontiers: testing, solving larger problems, code review, refactoring, documentation, and debugging. She emphasizes that while AI tools can help, they are not a silver bullet and require careful integration. The talk includes practical examples and references to resources like the GitClear report and Thoughtworks articles.

152 words

Critical Evaluation

The presentation offers a pragmatic and well-structured overview of the current state of AI assistance for software teams. Birgitta Böckeler leverages her extensive experience at Thoughtworks to provide insights that are grounded in real-world usage. The talk is particularly strong in its categorization of AI tooling and its discussion of practical considerations such as model choices, data privacy, and context orchestration. The speaker is honest about the limitations and hype fatigue, which adds credibility. However, the content is largely anecdotal and lacks rigorous empirical data; while she references reports like the GitClear study on refactoring vs. refuctoring, she does not delve into their findings in detail. The argumentation is coherent, but the lack of quantitative evidence weakens the scientific rigor. The sources cited are reputable (e.g., Thoughtworks, Martin Fowler, GitClear), but they are not systematically analyzed. The talk is more of an expert opinion than a systematic review. The title accurately reflects the content, and the presentation is well-organized with clear sections. Overall, it is a valuable resource for practitioners seeking to understand the landscape, but it should be complemented with more data-driven research.

184 words

Title / Content Match

The title accurately reflects the content, which provides an overview of AI assistance tools for software teams, focusing on coding assistants and their current state.

Quality & Reliability

8/10

The speaker is a Technical Principal at Thoughtworks with extensive hands-on experience in AI-assisted software delivery. The talk is based on practical experience and references several industry reports and articles. However, it is largely anecdotal and lacks rigorous empirical evidence.

Chapters

Cited Sources

Concurring Sources

Dissenting Sources

External References

Contribution & Novelties

The talk provides a current snapshot of AI assistance for software teams, synthesizing practical experience and industry resources. It offers a useful taxonomy of AI tooling and highlights key considerations for adoption, such as model quality and data privacy. The speaker’s emphasis on context orchestration and the limitations of current tools adds practical value.

Pour aller plus loin :

117 words

Radar Profile

The radar profile shows high scores in quantity and quality of information, with moderate technical depth and reliability. This indicates a well-informed talk with practical insights, but with room for more rigorous evidence.

Reliability 7/10