[{"data":1,"prerenderedAt":486},["ShallowReactive",2],{"/en-us/the-source/authors/sabrina-farmer/":3,"footer-en-us":31,"the-source-navigation-en-us":339,"the-source-newsletter-en-us":366,"sabrina-farmer-articles-list-authors-en-us":378,"sabrina-farmer-articles-list-en-us":409,"sabrina-farmer-page-categories-en-us":485},{"_path":4,"_dir":5,"_draft":6,"_partial":6,"_locale":7,"config":8,"seo":10,"content":12,"type":23,"slug":24,"_id":25,"_type":26,"title":11,"_source":27,"_file":28,"_stem":29,"_extension":30},"/en-us/the-source/authors/sabrina-farmer","authors",false,"",{"layout":9},"the-source",{"title":11},"Sabrina Farmer",[13,21],{"componentName":14,"type":14,"componentContent":15},"TheSourceAuthorHero",{"name":11,"role":16,"bio":17,"headshot":18},"Chief Technology Officer","Sabrina Farmer is the Chief Technology Officer at GitLab, where she leads software engineering, operations, and customer support teams to execute the company's technical vision and strategy and oversee the development and delivery of GitLab's products and services.\n\nPrior to GitLab, Sabrina spent nearly two decades at Google, where she most recently served as vice president of engineering, core infrastructure. During her tenure with Google, she was directly responsible for the reliability, performance, and efficiency of all of Google's billion-user products and infrastructure.\n\nA long-time advocate for women in technology, Farmer earned a B.S. in Computer Science at the University of New Orleans, where she established two scholarships to help level the playing field for inclusion and empowerment in technology.",{"altText":11,"config":19},{"src":20},"https://res.cloudinary.com/about-gitlab-com/image/upload/v1751463377/udmzbjjr5xrcrffdlphx.webp",{"componentName":22,"type":22},"TheSourceArticlesList","author","sabrina-farmer","content:en-us:the-source:authors:sabrina-farmer.yml","yaml","content","en-us/the-source/authors/sabrina-farmer.yml","en-us/the-source/authors/sabrina-farmer","yml",{"_path":32,"_dir":33,"_draft":6,"_partial":6,"_locale":7,"data":34,"_id":335,"_type":26,"title":336,"_source":27,"_file":337,"_stem":338,"_extension":30},"/shared/en-us/main-footer","en-us",{"text":35,"source":36,"edit":42,"contribute":47,"config":52,"items":57,"minimal":327},"Git is a trademark of Software Freedom Conservancy and our use of 'GitLab' is under license",{"text":37,"config":38},"View page source",{"href":39,"dataGaName":40,"dataGaLocation":41},"https://gitlab.com/gitlab-com/marketing/digital-experience/about-gitlab-com/","page source","footer",{"text":43,"config":44},"Edit this page",{"href":45,"dataGaName":46,"dataGaLocation":41},"https://gitlab.com/gitlab-com/marketing/digital-experience/about-gitlab-com/-/blob/main/content/","web ide",{"text":48,"config":49},"Please contribute",{"href":50,"dataGaName":51,"dataGaLocation":41},"https://gitlab.com/gitlab-com/marketing/digital-experience/about-gitlab-com/-/blob/main/CONTRIBUTING.md/","please contribute",{"twitter":53,"facebook":54,"youtube":55,"linkedin":56},"https://twitter.com/gitlab","https://www.facebook.com/gitlab","https://www.youtube.com/channel/UCnMGQ8QHMAnVIsI3xJrihhg","https://www.linkedin.com/company/gitlab-com",[58,85,158,226,288],{"title":59,"links":60,"subMenu":66},"Platform",[61],{"text":62,"config":63},"DevSecOps platform",{"href":64,"dataGaName":65,"dataGaLocation":41},"/platform/","devsecops platform",[67],{"title":68,"links":69},"Pricing",[70,75,80],{"text":71,"config":72},"View plans",{"href":73,"dataGaName":74,"dataGaLocation":41},"/pricing/","view plans",{"text":76,"config":77},"Why Premium?",{"href":78,"dataGaName":79,"dataGaLocation":41},"/pricing/premium/","why premium",{"text":81,"config":82},"Why Ultimate?",{"href":83,"dataGaName":84,"dataGaLocation":41},"/pricing/ultimate/","why ultimate",{"title":86,"links":87},"Solutions",[88,93,98,103,108,113,118,123,128,133,138,143,148,153],{"text":89,"config":90},"Digital transformation",{"href":91,"dataGaName":92,"dataGaLocation":41},"/topics/digital-transformation/","digital transformation",{"text":94,"config":95},"Security & Compliance",{"href":96,"dataGaName":97,"dataGaLocation":41},"/solutions/security-compliance/","security & compliance",{"text":99,"config":100},"Automated software delivery",{"href":101,"dataGaName":102,"dataGaLocation":41},"/solutions/delivery-automation/","automated software delivery",{"text":104,"config":105},"Agile development",{"href":106,"dataGaName":107,"dataGaLocation":41},"/solutions/agile-delivery/","agile delivery",{"text":109,"config":110},"Cloud transformation",{"href":111,"dataGaName":112,"dataGaLocation":41},"/topics/cloud-native/","cloud transformation",{"text":114,"config":115},"SCM",{"href":116,"dataGaName":117,"dataGaLocation":41},"/solutions/source-code-management/","source code management",{"text":119,"config":120},"CI/CD",{"href":121,"dataGaName":122,"dataGaLocation":41},"/solutions/continuous-integration/","continuous integration & delivery",{"text":124,"config":125},"Value stream management",{"href":126,"dataGaName":127,"dataGaLocation":41},"/solutions/value-stream-management/","value stream management",{"text":129,"config":130},"GitOps",{"href":131,"dataGaName":132,"dataGaLocation":41},"/solutions/gitops/","gitops",{"text":134,"config":135},"Enterprise",{"href":136,"dataGaName":137,"dataGaLocation":41},"/enterprise/","enterprise",{"text":139,"config":140},"Small business",{"href":141,"dataGaName":142,"dataGaLocation":41},"/small-business/","small business",{"text":144,"config":145},"Public sector",{"href":146,"dataGaName":147,"dataGaLocation":41},"/solutions/public-sector/","public sector",{"text":149,"config":150},"Education",{"href":151,"dataGaName":152,"dataGaLocation":41},"/solutions/education/","education",{"text":154,"config":155},"Financial services",{"href":156,"dataGaName":157,"dataGaLocation":41},"/solutions/finance/","financial services",{"title":159,"links":160},"Resources",[161,166,171,176,181,186,191,196,201,206,211,216,221],{"text":162,"config":163},"Install",{"href":164,"dataGaName":165,"dataGaLocation":41},"/install/","install",{"text":167,"config":168},"Quick start guides",{"href":169,"dataGaName":170,"dataGaLocation":41},"/get-started/","quick setup checklists",{"text":172,"config":173},"Learn",{"href":174,"dataGaName":175,"dataGaLocation":41},"https://university.gitlab.com/","learn",{"text":177,"config":178},"Product documentation",{"href":179,"dataGaName":180,"dataGaLocation":41},"https://docs.gitlab.com/","docs",{"text":182,"config":183},"Blog",{"href":184,"dataGaName":185,"dataGaLocation":41},"/blog/","blog",{"text":187,"config":188},"Customer success stories",{"href":189,"dataGaName":190,"dataGaLocation":41},"/customers/","customer success stories",{"text":192,"config":193},"Remote",{"href":194,"dataGaName":195,"dataGaLocation":41},"https://handbook.gitlab.com/handbook/company/culture/all-remote/","remote",{"text":197,"config":198},"GitLab Services",{"href":199,"dataGaName":200,"dataGaLocation":41},"/services/","services",{"text":202,"config":203},"TeamOps",{"href":204,"dataGaName":205,"dataGaLocation":41},"/teamops/","teamops",{"text":207,"config":208},"Community",{"href":209,"dataGaName":210,"dataGaLocation":41},"/community/","community",{"text":212,"config":213},"Forum",{"href":214,"dataGaName":215,"dataGaLocation":41},"https://forum.gitlab.com/","forum",{"text":217,"config":218},"Events",{"href":219,"dataGaName":220,"dataGaLocation":41},"/events/","events",{"text":222,"config":223},"Partners",{"href":224,"dataGaName":225,"dataGaLocation":41},"/partners/","partners",{"title":227,"links":228},"Company",[229,234,239,244,249,254,259,263,268,273,278,283],{"text":230,"config":231},"About",{"href":232,"dataGaName":233,"dataGaLocation":41},"/company/","company",{"text":235,"config":236},"Jobs",{"href":237,"dataGaName":238,"dataGaLocation":41},"/jobs/","jobs",{"text":240,"config":241},"Leadership",{"href":242,"dataGaName":243,"dataGaLocation":41},"/company/team/e-group/","leadership",{"text":245,"config":246},"Team",{"href":247,"dataGaName":248,"dataGaLocation":41},"/company/team/","team",{"text":250,"config":251},"Handbook",{"href":252,"dataGaName":253,"dataGaLocation":41},"https://handbook.gitlab.com/","handbook",{"text":255,"config":256},"Investor relations",{"href":257,"dataGaName":258,"dataGaLocation":41},"https://ir.gitlab.com/","investor relations",{"text":260,"config":261},"Sustainability",{"href":262,"dataGaName":260,"dataGaLocation":41},"/sustainability/",{"text":264,"config":265},"Diversity, inclusion and belonging (DIB)",{"href":266,"dataGaName":267,"dataGaLocation":41},"/diversity-inclusion-belonging/","Diversity, inclusion and belonging",{"text":269,"config":270},"Trust Center",{"href":271,"dataGaName":272,"dataGaLocation":41},"/security/","trust center",{"text":274,"config":275},"Newsletter",{"href":276,"dataGaName":277,"dataGaLocation":41},"/company/contact/","newsletter",{"text":279,"config":280},"Press",{"href":281,"dataGaName":282,"dataGaLocation":41},"/press/","press",{"text":284,"config":285},"Modern Slavery Transparency Statement",{"href":286,"dataGaName":287,"dataGaLocation":41},"https://handbook.gitlab.com/handbook/legal/modern-slavery-act-transparency-statement/","modern slavery transparency statement",{"title":289,"links":290},"Contact Us",[291,296,301,306,311,316,321],{"text":292,"config":293},"Contact an expert",{"href":294,"dataGaName":295,"dataGaLocation":41},"/sales/","sales",{"text":297,"config":298},"Get help",{"href":299,"dataGaName":300,"dataGaLocation":41},"/support/","get help",{"text":302,"config":303},"Customer portal",{"href":304,"dataGaName":305,"dataGaLocation":41},"https://customers.gitlab.com/customers/sign_in/","customer portal",{"text":307,"config":308},"Status",{"href":309,"dataGaName":310,"dataGaLocation":41},"https://status.gitlab.com/","status",{"text":312,"config":313},"Terms of use",{"href":314,"dataGaName":315,"dataGaLocation":41},"/terms/","terms of use",{"text":317,"config":318},"Privacy statement",{"href":319,"dataGaName":320,"dataGaLocation":41},"/privacy/","privacy statement",{"text":322,"config":323},"Cookie preferences",{"dataGaName":324,"dataGaLocation":41,"id":325,"isOneTrustButton":326},"cookie preferences","ot-sdk-btn",true,{"items":328},[329,331,333],{"text":312,"config":330},{"href":314,"dataGaName":315,"dataGaLocation":41},{"text":317,"config":332},{"href":319,"dataGaName":320,"dataGaLocation":41},{"text":322,"config":334},{"dataGaName":324,"dataGaLocation":41,"id":325,"isOneTrustButton":326},"content:shared:en-us:main-footer.yml","Main Footer","shared/en-us/main-footer.yml","shared/en-us/main-footer",{"_path":340,"_dir":9,"_draft":6,"_partial":6,"_locale":7,"logo":341,"subscribeLink":346,"navItems":350,"_id":362,"_type":26,"title":363,"_source":27,"_file":364,"_stem":365,"_extension":30},"/shared/en-us/the-source/navigation",{"altText":342,"config":343},"the source logo",{"src":344,"href":345},"https://res.cloudinary.com/about-gitlab-com/image/upload/v1750191004/t7wz1klfb2kxkezksv9t.svg","/the-source/",{"text":347,"config":348},"Subscribe",{"href":349},"#subscribe",[351,355,358],{"text":352,"config":353},"Artificial Intelligence",{"href":354},"/the-source/ai/",{"text":94,"config":356},{"href":357},"/the-source/security/",{"text":359,"config":360},"Platform & Infrastructure",{"href":361},"/the-source/platform/","content:shared:en-us:the-source:navigation.yml","Navigation","shared/en-us/the-source/navigation.yml","shared/en-us/the-source/navigation",{"_path":367,"_dir":9,"_draft":6,"_partial":6,"_locale":7,"title":368,"description":369,"submitMessage":370,"formData":371,"_id":375,"_type":26,"_source":27,"_file":376,"_stem":377,"_extension":30},"/shared/en-us/the-source/newsletter","The Source Newsletter","Stay updated with insights for the future of software development.","You have successfully signed up for The Source’s newsletter.",{"config":372},{"formId":373,"formName":374,"hideRequiredLabel":326},1077,"thesourcenewsletter","content:shared:en-us:the-source:newsletter.yml","shared/en-us/the-source/newsletter.yml","shared/en-us/the-source/newsletter",{"amanda-rueda":379,"andre-michael-braun":380,"andrew-haschka":381,"ayoub-fandi":382,"bob-stevens":383,"brian-wald":384,"bryan-ross":385,"chandler-gibbons":386,"dave-steer":387,"ddesanto":388,"derek-debellis":389,"emilio-salvador":390,"erika-feldman":391,"george-kichukov":392,"gitlab":393,"grant-hickman":394,"haim-snir":395,"iganbaruch":396,"jlongo":397,"joel-krooswyk":398,"josh-lemos":399,"julie-griffin":400,"kristina-weis":401,"lee-faus":402,"ncregan":403,"rschulman":404,"sabrina-farmer":11,"sandra-gittlen":405,"sharon-gaudin":406,"stephen-walters":407,"taylor-mccaslin":408},"Amanda Rueda","Andre Michael Braun","Andrew Haschka","Ayoub Fandi","Bob Stevens","Brian Wald","Bryan Ross","Chandler Gibbons","Dave Steer","David DeSanto","Derek DeBellis","Emilio Salvador","Erika Feldman","George Kichukov","GitLab","Grant Hickman","Haim Snir","Itzik Gan Baruch","Joseph Longo","Joel Krooswyk","Josh Lemos","Julie Griffin","Kristina Weis","Lee Faus","Niall Cregan","Robin Schulman","Sandra Gittlen","Sharon Gaudin","Stephen Walters","Taylor McCaslin",{"allArticles":410,"visibleArticles":484,"showAllBtn":326},[411,451],{"_path":412,"_dir":413,"_draft":6,"_partial":6,"_locale":7,"slug":414,"type":415,"category":413,"config":416,"seo":420,"content":424,"_id":448,"_type":26,"title":421,"_source":27,"_file":449,"_stem":450,"_extension":30,"description":422,"date":425,"timeToRead":426,"keyTakeaways":427,"articleBody":431,"faq":432,"heroImage":423},"/en-us/the-source/ai/three-ways-to-operationalize-ai-for-engineering-teams","ai","three-ways-to-operationalize-ai-for-engineering-teams","article",{"layout":9,"template":417,"featured":326,"articleType":418,"author":24,"gatedAsset":419,"isHighlighted":6,"authorName":11},"TheSourceArticle","Regular","source-lp-how-to-get-started-using-ai-in-software-development",{"title":421,"description":422,"ogImage":423},"Three ways to operationalize AI for engineering teams","Discover three actionable frameworks for engineering leaders to implement AI strategically, drive measurable ROI, and overcome adoption barriers.","https://res.cloudinary.com/about-gitlab-com/image/upload/v1751908411/i1mwfh3egxgbx5ijkowi.png",{"title":421,"description":422,"date":425,"timeToRead":426,"keyTakeaways":427,"articleBody":431,"faq":432,"heroImage":423},"2025-07-08","4 min read",[428,429,430],"AI adoption succeeds when positioned as a collaborative development partner — similar to pair programming — with specific applications like enhanced debugging, solution architecture, and code quality assurance rather than a replacement for engineers.","Strategic AI implementation requires role-specific applications with clear ROI targets, seamless workflow integration that minimizes friction, and structured feedback loops that connect AI initiatives directly to business outcomes.","Incremental implementation victories, rather than wholesale transformation, drive successful AI adoption — with success measured through problem-solving effectiveness and business impact instead of traditional productivity metrics.","Technical leaders face mounting pressure to adopt AI tools, but many struggle to move beyond experimentation to systematic implementation that delivers measurable ROI. While AI's potential for software development is clear, the path to operationalization remains challenging.\n\n[GitLab research](https://about.gitlab.com/developer-survey/2024/ai/) reveals that approximately half of organizations are still in the evaluation and exploration stage of AI maturity. These teams recognize AI's potential but haven't crystallized their implementation strategy, a common challenge I've observed when speaking with engineering executives.\n\n## Breaking through implementation barriers\n\nTwo critical obstacles stand in the way of successful AI adoption. First is the fear that AI will replace human engineers — a legitimate concern requiring transparent communication from leadership. Second, it is important to determine where to begin implementing AI when many engineers see limited value in disrupting established workflows.\n\nTechnical leaders must reframe AI’s value proposition by connecting AI capabilities directly to business outcomes. [Success metrics](https://about.gitlab.com/the-source/ai/4-steps-for-measuring-the-impact-of-ai/) should focus on problem-solving effectiveness and business impact rather than code volume or traditional individual productivity measures.\n\nRather than viewing AI as a threat to jobs, help your teams consider it through the lens of established collaborative practices like pair programming. This familiar framework provides clear entry points for AI integration:\n\n* **Enhanced debugging partner**: AI functions as a sophisticated \"[rubber duck](https://rubberduckdebugging.com/)\" that not only listens but responds with actionable insights\n* **Solution architect**: AI can generate multiple implementation approaches to complex problems within seconds\n* **Code quality guardian**: AI can help teams identify optimization opportunities and vulnerabilities before human review\n\nWhen positioned as an augmentation layer that eliminates repetitive tasks and amplifies human creativity, AI becomes an enabler rather than a threat.\n\n## A three-step implementation framework for technical leaders\n\nTo integrate AI into team workflows, leadership must first establish the context and then take a top-down approach to implementation. Specifically, leaders must define how teams will use AI, establish clear processes, and provide the necessary resources and support. Rather than overhauling your team's existing workflows entirely, apply AI to specific tasks or stages of the development process. This iterative approach allows teams to learn, adapt, and build confidence in AI over time.\n\n### 1. Define role-specific AI applications with clear ROI\n\nInstead of vague directives, specify exactly how different roles will leverage AI:\n\n* **Developers**: Ensure a consistent and thorough initial analysis and mandate AI-powered first code reviews and security scans before your human review. Leveraging AI first to analyze code for potential bugs, vulnerabilities, and performance issues can provide developers with actionable insights for remediation, while also creating learning moments.\n* **Quality assurance (QA) engineers**: Use AI to generate the first test for new code and analyze test results, freeing developers to focus on more complex testing scenarios and critical issues. Editing a proposed test is typically easier than generating it from scratch.\n* **Operations teams**: Implement AI to automate repetitive operational tasks such as deployments and infrastructure management and monitoring to free up operations teams' time for more strategic work.\n* **Team leads**: Leverage AI to assist with project planning, backlog prioritization, resource allocation, initial triage, and progress tracking, providing team leads with real-time insights into project health and potential risks.\n* **Product managers**: Use AI to analyze and summarize customer verticals, market trends, customer forums, and overall customer sentiment.\n\n### 2. Integrate AI seamlessly into existing workflows\n\nSelect AI solutions that seamlessly integrate into your existing development environment to avoid additional burdens on your developers. To avoid decision fatigue, develop clear guidelines for when and how to use AI tools, including:\n\n* When to rely on AI-generated suggestions\n* How to critically evaluate AI recommendations\n* What feedback mechanisms exist for improving AI outputs\n\n### 3. Create feedback loops and measure business impact\n\nEstablish structured communication channels for engineers to share AI wins and challenges. Create internal communities of practice around AI integration to accelerate knowledge sharing. Encourage developers to interact with the AI, provide feedback on generated code, refine test cases, and actively participate in the collaborative process.\n\nAfter implementation, quantify and communicate the business impact to executive stakeholders. It’s important to position AI not as experimental technology but as a strategic lever for competitive advantage and engineering excellence.\n\n## Moving beyond experimentation\n\nThe key to successful AI operationalization is targeted implementation with clear business objectives. By defining role-specific applications, creating seamless integration points, and establishing feedback mechanisms, engineering leaders can transform AI from an interesting curiosity to a foundational productivity multiplier.\n\nSuccess will not come from wholesale workflow transformation but through incremental victories demonstrating tangible value. With this structured approach, technical leaders can unlock AI's true potential while ensuring their teams feel empowered rather than threatened by this technological evolution.",[433,436,439,442,445],{"header":434,"content":435},"What percentage of organizations are still evaluating AI implementation?","Approximately half of organizations remain in the evaluation and exploration stage of AI maturity. These teams recognize AI's potential but haven't crystallized their implementation strategy, creating a common challenge for engineering executives moving beyond experimentation.",{"header":437,"content":438},"How should engineering leaders position AI to overcome adoption resistance?","Leaders should reframe AI as a collaborative development partner similar to pair programming rather than a replacement. Position AI as an enhanced debugging partner, solution architect, and code quality guardian that eliminates repetitive tasks while amplifying human creativity.",{"header":440,"content":441},"What are the three key steps for implementing AI in engineering workflows?","First, define role-specific AI applications with clear ROI for developers, QA engineers, operations teams, team leads, and product managers. Second, integrate AI seamlessly into existing development environments. Third, create feedback loops and measure business impact through structured communication channels.",{"header":443,"content":444},"How should AI success be measured in engineering teams?","Success metrics should focus on problem-solving effectiveness and business impact rather than code volume or traditional productivity measures. Quantify business impact for executive stakeholders and position AI as a strategic lever for competitive advantage and engineering excellence.",{"header":446,"content":447},"What AI applications work best for different engineering roles?","Developers use AI for code reviews and security scans. QA engineers leverage AI for test generation and result analysis. Operations teams implement AI for deployments and infrastructure monitoring. Team leads use AI for project planning and progress tracking. Product managers apply AI for customer sentiment analysis.","content:en-us:the-source:ai:three-ways-to-operationalize-ai-for-engineering-teams:index.yml","en-us/the-source/ai/three-ways-to-operationalize-ai-for-engineering-teams/index.yml","en-us/the-source/ai/three-ways-to-operationalize-ai-for-engineering-teams/index",{"_path":452,"_dir":413,"_draft":6,"_partial":6,"_locale":7,"config":453,"seo":455,"content":459,"type":415,"slug":480,"category":413,"_id":481,"_type":26,"title":456,"_source":27,"_file":482,"_stem":483,"_extension":30,"date":460,"description":457,"timeToRead":461,"heroImage":458,"keyTakeaways":462,"articleBody":466,"faq":467},"/en-us/the-source/ai/three-challenges-impacting-your-teams-ai-productivity-gains",{"layout":9,"template":417,"articleType":418,"author":24,"featured":326,"gatedAsset":454,"isHighlighted":6,"authorName":11},"source-lp-ai-guide-for-enterprise-leaders-building-the-right-approach",{"title":456,"description":457,"ogImage":458},"Three challenges impacting your team’s AI productivity gains","AI is becoming a critical part of software development — but there are growing pains. Learn more about common roadblocks and how to address them.","https://res.cloudinary.com/about-gitlab-com/image/upload/v1751464418/sekku5gned7o9tct0jze.png",{"title":456,"date":460,"description":457,"timeToRead":461,"heroImage":458,"keyTakeaways":462,"articleBody":466,"faq":467},"2025-01-23","5 min read",[463,464,465],"AI can increase software development productivity by automating tasks, identifying insights from large datasets, and reducing time spent on repetitive tasks. However, there are challenges to achieving these productivity gains.","Organizations may face challenges such as an AI training gap, toolchain sprawl, and appropriately defining productivity metrics. Addressing these can help ensure the effective utilization of AI in software development.","To evaluate AI's effectiveness, organizations should measure ROI based on user adoption, time to market, revenue, and customer satisfaction metrics. Evaluation of the right metrics can help organizations better understand AI's impact on business outcomes.","Software development is at a turning point. AI promises to transform development workflows, but many organizations are discovering that integrating AI effectively requires more than just adopting new tools. [A GitLab research study](https://about.gitlab.com/developer-survey/) revealed that while executives are confident about AI adoption, 25% of individual contributors report their organizations aren’t providing adequate training and resources to help them use AI.\n\nAI can help teams tackle increasingly complex challenges, from code generation and security vulnerability detection to automated testing and project management. When implemented thoughtfully, AI allows developers to focus on innovation rather than repetition, leading to improved code quality. More importantly, AI’s ability to analyze vast datasets of code, builds, and deployments helps teams make informed decisions that accelerate delivery while reducing risks.\n\nHowever, as AI technology becomes more integrated into software development processes, organizations encounter three key challenges that can hinder these potential productivity gains.\n\n## 1. The AI training gap\nThe executive/developer perception gap isn’t surprising: While executives focus on AI’s strategic potential, development teams face the day-to-day reality of integrating these tools into their workflows. The disconnect often stems from organizations viewing AI as a potential replacement for software engineers, rather than a tool that enables more creative and strategic human-centered work. Software leaders should supplement their investments in AI with investments in training and development resources that allow software development teams to build momentum and motivation over time.\n\nIt’s important to call out here that your teams will need a grace period to determine how AI best fits their processes. Initially, productivity may decline as they adjust to new workflows. However, your teams will build trust in their new tools by testing how AI can best fit into their day-to-day workflows and see better results.\n\n## 2. AI-powered toolchain sprawl\nOne major factor that can detract from developer experience and impact overall productivity is [toolchain sprawl](https://about.gitlab.com//the-source/platform/devops-teams-want-to-shake-off-diy-toolchains-a-platform-is-the-answer/), or having multiple point solutions across the software development process. GitLab’s research found that two-thirds of DevSecOps professionals want to consolidate their toolchain, with many citing negative impacts on developer experience caused by context switching between tools.\n\nToolchain sprawl has additional drawbacks, such as adding cost and complexity, creating silos, and making it more challenging to standardize processes across teams. It also creates security concerns due to expanding attack surfaces and unnecessary handoff points. AI-powered point solutions compound these issues. In fact, GitLab’s research found that respondents whose organizations are currently using AI were more likely to want to consolidate their toolchains than those not using AI - even though there wasn’t a significant difference between the two groups in the number of tools respondents reported using.\n\nRather than attempting to integrate AI into unwieldy, complex toolchains, adopt consistent, strategic best practices that [minimize your teams’ context switching and cognitive load](https://about.gitlab.com/the-source/ai/devops-leaders-fix-this-productivity-blocker-before-adding-ai/) while reducing your organization’s total cost of ownership. Before incorporating new AI development tools, [evaluate your existing toolchains](https://about.gitlab.com/the-source/ai/overcome-ai-sprawl-with-a-value-stream-management-approach/) to determine areas where you can streamline or eliminate disparate tools to avoid the strain of integrating excess tools with AI-powered solutions.\n\n## 3. Unclear productivity metrics\nDeveloper productivity is a top concern for the C-suite. While measuring developer productivity has always been difficult, [AI has compounded the challenge](https://about.gitlab.com/the-source/ai/4-steps-for-measuring-the-impact-of-ai/). You might agree that measuring developer productivity can help business growth, but most leaders aren’t effectively measuring productivity against business priorities. GitLab’s research revealed that less than half (42%) of C-level executives currently measure developer productivity within their organization and are happy with their approach.\n\nMany organizations struggle to quantify the impact of AI-powered tools on developer productivity or other real-world business outcomes. Traditional metrics, such as lines of code, code commits, or task completion, are often insufficient when assessing development’s impact on a business’s bottom line.\n\nThe best approach to modernizing measurement practices begins with consolidating quantitative data from throughout the software development lifecycle with insights from software developers on how AI is supporting or hindering their daily work.\n\n## Making AI work for your teams\nSuccessfully implementing AI in software development requires closing the gap between executive expectations and developer realities. Start where your team feels the most friction today— whether that’s providing proper training, consolidating toolchains, or rethinking traditional productivity metrics. Taking action now allows your teams to realize meaningful productivity gains, rather than just adding new tools.",[468,471,474,477],{"header":469,"content":470},"How is the gap between executive expectations and developer experience affecting AI adoption in software development?","While executives remain optimistic about AI’s strategic potential, many developers face challenges integrating these tools into daily workflows. This disconnect can result from a lack of training and support, with some organizations viewing AI as a replacement for developers rather than an enabler of more meaningful work. Addressing this gap requires investments in developer education and a grace period to adapt to new AI-driven workflows.",{"header":472,"content":473},"Why is toolchain sprawl a problem when implementing AI in software development?","Toolchain sprawl, or using multiple point solutions across development processes, can negatively impact developer experience by increasing context switching and complexity. AI-powered tools can worsen this issue if introduced into already fragmented toolchains, creating additional silos and security risks. Streamlining tools and adopting integrated solutions can reduce friction, improve productivity, and lower total cost of ownership.",{"header":475,"content":476},"What makes measuring AI-driven productivity difficult for organizations?","Many organizations struggle to quantify the value of AI tools using traditional metrics like lines of code or task completion. These measures often fall short in reflecting how AI impacts overall business outcomes. A more effective approach combines lifecycle-wide quantitative data with qualitative insights from developers to understand how AI supports or hinders day-to-day work.",{"header":478,"content":479},"What steps can organizations take to improve the impact of AI on developer productivity?","Organizations can begin by addressing the most pressing friction points for their teams. This might include offering better AI training, simplifying complex toolchains, or modernizing productivity metrics. A strategic, developer-first approach ensures AI is integrated in a way that enhances rather than complicates development workflows.","three-challenges-impacting-your-teams-ai-productivity-gains","content:en-us:the-source:ai:three-challenges-impacting-your-teams-ai-productivity-gains:index.yml","en-us/the-source/ai/three-challenges-impacting-your-teams-ai-productivity-gains/index.yml","en-us/the-source/ai/three-challenges-impacting-your-teams-ai-productivity-gains/index",[411,451],{"ai":352,"platform":359,"security":94},1753733242065]