Insights
Points of view from the work: what AI is really changing in software delivery, and what engineering leaders can do about it.
Latest
Engineering quality
AI-written code and the security gap scanners miss
AI-generated code often passes a quick review and a generic scan, and still carries real risk into production. What the research shows, why it happens, and how to close the gap.
2 min read
Strategy
Where do you really stand with AI? A maturity framework for software delivery
License counts and usage dashboards do not tell you how mature your AI adoption is. A five-level, six-dimension framework does, and it turns into a roadmap you can fund and track.
3 min read
Commercials
When AI does more of the work, what exactly are you paying for?
Rate cards and headcount pricing assumed that effort meant output. AI has broken that link. A look at how software delivery should be priced now, for buyers and for services firms.
3 min read
People
What a forward-deployed engineer really is, and how to tell
Every vendor and candidate now claims to be an FDE. Here is what the role actually involves, how it differs from a specialist in a traditional relay, and how to assess it.
3 min read
Frameworks
The models behind how we work.
The YuktiMedha AI-SDLC maturity framework
Five levels. Six dimensions. One roadmap.
- Level 1Exploring
- Level 2Adopting
- Level 3Integrated
- Level 4Optimized
- Level 5AI-native
Assessed across
- Strategy and value
- Process and delivery
- People and roles
- Engineering quality
- Governance and risk
- Measures and economics
AI-SDLC maturity framework
Five levels and six dimensions for assessing where you stand and building a roadmap you can fund.
Read the frameworkThe Forward Loop
Forward-deployed engineering: a small senior pod that shapes, builds, verifies, ships and learns in short loops.
See our approachResearch we cite
The evidence, with sources.
Every figure we use links to its source. Here is the research behind our points of view, and our take on each.
See all research notesGoogle DORA2024
Every 25% increase in AI adoption was associated with an estimated 7.2% drop in delivery stability and a 1.5% drop in throughput.
METR2025
Experienced developers took 19% longer to complete real tasks with AI tools, while believing they had been about 20% faster.
Veracode2025
45% of AI-generated code samples failed security tests, across more than 100 language models.
GitClear2025
Duplicated code blocks increased eightfold during 2024, while moved (refactored) code fell by almost 40%.
Want to talk through any of these?
Start with a short conversation, or follow new articles through our RSS feed.