Incorporating AI Into Your SDLC: Leveraging AI tooling across the phases of your software development lifecycle

Presented at All Things Open 2025
Presented by Brent Laster - Tech Skills Transformations

Title: Incorporating AI Into Your SDLC: Leveraging AI tooling across the phases of your software development lifecycle
Abstract: Just as CI/CD and other revolutions in DevOps have changed the landscape of the software development lifecycle (SDLC), so Generative AI is now changing it again. Gen AI has the potential to simplify, clarify, and lessen the cycles required across multiple phases of the SDLC.

In this session with author, trainer, and experienced DevOps director Brent Laster, we'll survey the ways that today's AI assistants and tools can be incorporated across your SDLC phases including planning, development, testing, documentation, maintaining, etc. There are multiple ways the existing tools can help us beyond just the standard day-to-day coding and, like other changes that have happened over the years, teams need to be aware of, and thinking about how to incorporate AI into their processes to stay relevant and up-to-date.

Find more info about All Things Open:
On the web: https://www.allthingsopen.org/
Twitter: https://twitter.com/AllThingsOpen
LinkedIn: https://www.linkedin.com/company/all-things-open/
Instagram: https://www.instagram.com/allthingsopen/
Facebook: https://www.facebook.com/AllThingsOpen
Mastodon: https://mastodon.social/@allthingsopen
Threads: https://www.threads.net/@allthingsopen
Bluesky: https://bsky.app/profile/allthingsopen.bsky.social
YouTube: https://www.youtube.com/@allthingsopen
2025 conference: https://2025.allthingsopen.org/

1. © 2024 Brent C. Laster &techupskills.com | techskillstransformations.com
© 2025 Brent C. Laster &
@techupskills
2
Brent Laster
© 2025 Brent C. Laster & Tech Skills Transformations LLC

All rights reserved
Incorporating AI into your SDLC
Leveraging AI tooling across the phases of your software development lifecycle
Presented by
Tech Skills Transformations LLC

2. © 2024 Brent C. Laster &techupskills.com | techskillstransformations.com
© 2025 Brent C. Laster &
@techupskills
3
What we're going to be talking about...
•Roles for using AI
•Where are we today?
•Basic Tenets
•Using AI in particular phases
▪Speeding up planning with RAG and prototyping
▪13 use cases for AI in development
▪Testing by direction, exploration and best practices
▪Using agents to resolve issues
▪Getting the documentation monkey off your back with AI
▪Streamlining maintenance
▪Using AI to help with deployment

3. © 2024 Brent C. Laster &techupskills.com | techskillstransformations.com
© 2025 Brent C. Laster &
@techupskills
4
About me

•Founder, Tech Skills
Transformations LLC
•https://getskillsnow.com
•[email protected]
❑LinkedIn: brentlaster
❑X: @BrentCLaster
❑Bluesky:
brentclaster.bsky.social
❑GitHub: brentlaster
Long career in corporate:
•Principal Dev
•Manager/Senior Manager
•Director

4. © 2024 Brent C. Laster &techupskills.com | techskillstransformations.com
© 2025 Brent C. Laster &
@techupskills
5
Upcoming AI Training (nofluffjuststuff.com & oreilly.com

5. © 2024 Brent C. Laster &techupskills.com | techskillstransformations.com
© 2025 Brent C. Laster &
@techupskills
6
Enterprise AI Accelerator Training

6. © 2024 Brent C. Laster &techupskills.com | techskillstransformations.com
© 2025 Brent C. Laster &
@techupskills
7
Types of AI Assistance for Coding and Related Tasks
•Single-panel interfaces
•Extensions that add AI code completions
and chat
•AI Coding IDEs and CLIs
▪AI-first design
▪Tend to focus more fully on project-
wide insights and application
building
▪Think “collaborate like Copilot and
handle complex tasks like an agent”
•Real-time AI assistants
▪Offer answers based in current
content
▪Synthesize info from multiple
sources (cited) into concise, natural
language answers
▪Uses LLM on back end
▪Great at real-time info retrieval
•Users can leverage these in three modes:
▪Acceleration
▪Exploration
▪Directive

7. © 2024 Brent C. Laster &techupskills.com | techskillstransformations.com
© 2025 Brent C. Laster &
@techupskills
8
Roles and Uses
•Accelerators/Acceleration - code completion
- working in the flow and AI provides
completion suggestions (usually in the editor)
•Explorers/Exploration - using the chat
interface to ask questions or find out
information
•Directors/Direction - giving the AI direct and
specific instructions to create something via
the chat interface or in the editor (via
comments)
•All of these can lead to increased developer
productivity across the phases of the SDLC.
source: "https://www.mckinsey.com/capabilities/mckinsey-digital/our-insights/unleashing-developer-productivity-with-generative-ai

8. © 2024 Brent C. Laster &techupskills.com | techskillstransformations.com
© 2025 Brent C. Laster &
@techupskills
9
The Software Development Lifecycle (SDLC)
•Structured process used to create high-quality
software efficiently and consistently
•Outlines and defines the steps to plan, develop, test,
and deploy
•Typical phases include
▪Planning
▪Analysis
▪Design
▪Implementation
▪Testing and integration
▪Maintenance
•Defined phases may vary across
organizations/projects

9. © 2024 Brent C. Laster &techupskills.com | techskillstransformations.com
© 2025 Brent C. Laster &
@techupskills
10
Where are we today with AI tool use?

10. © 2024 Brent C. Laster &techupskills.com | techskillstransformations.com
© 2025 Brent C. Laster &
@techupskills
11
|
Using AI within the SDLC

11. © 2024 Brent C. Laster &techupskills.com | techskillstransformations.com
© 2025 Brent C. Laster &
@techupskills
12
Basic Tenets
•In terms of skill level, treat your AI
like an intern or someone new to
the team
▪AI-generated content MUST be
subject to the same
validation/testing/review you would
use for non AI-generated content
•Selecting the correct context and
using good prompting strategies are
keys
•Invest time in learning and using

12. © 2024 Brent C. Laster &techupskills.com | techskillstransformations.com
© 2025 Brent C. Laster &
@techupskills
13
|
Planning

13. © 2024 Brent C. Laster &techupskills.com | techskillstransformations.com
© 2025 Brent C. Laster &
@techupskills
14
Planning: Prototyping
•For brainstorming/considering requirements
•Can quickly draft pseudo code in comments
•AI can generate "draft" code
•Doesn't have to be perfect, just enough for
estimation/planning

14. © 2024 Brent C. Laster &techupskills.com | techskillstransformations.com
© 2025 Brent C. Laster &
@techupskills
15
Planning: Leverage Your Data
•Index your code and docs for ability to search by related meaning (semantic search)
•Incorporate indices with AI to get the best of both (useful AI responses grounded in
your data) (RAG)
•Ask questions of your code to simplify estimates, avoid rework, accelerate planning

15. © 2024 Brent C. Laster &techupskills.com | techskillstransformations.com
© 2025 Brent C. Laster &
@techupskills
16
Prepping your data for searching and use with AI
Embedding
Model
(convert to
numeric form
with context)
Data Store
Document
Embeddings
•Your data is parsed and stored with information about other data it's related to in a data store
Retrieve/Parse
Code/Docs
Code
Repositories (or
Documents /
Knowledge Base)

16. © 2024 Brent C. Laster &techupskills.com | techskillstransformations.com
© 2025 Brent C. Laster &
@techupskills
17
Integrating your data searches with AI (RAG)
Embedding
Model
Data Store
LLM
User
Interface
Prompt
Document
Embeddings
embedded
query Prompt + enhanced
context
response (generative)User Query and Response Generation
Prompt
Prompt
Original prompt +
matching "docs" (aka
"enhanced context")
LLM Response
-----------
------------------
---------
-------------
---------------------
•For queries/prompts, application gathers
results (most relevant ones) from the
vector database with your data
•Adds results to your regular LLM
query/prompt
•Asks the LLM to answer based on the
augmented/enriched query/prompt
•NOTE: Items returned via RAG search are
existing items from the data store, not
generated content

17. © 2024 Brent C. Laster &techupskills.com | techskillstransformations.com
© 2025 Brent C. Laster &
@techupskills
18RAG Example

18. © 2024 Brent C. Laster &techupskills.com | techskillstransformations.com
© 2025 Brent C. Laster &
@techupskills
19
|
Development

19. © 2024 Brent C. Laster &techupskills.com | techskillstransformations.com
© 2025 Brent C. Laster &
@techupskills
20
Ways Copilot Can Help
•Understanding a New Codebase
▪Explain code
▪Identify key files, entry points,
main modules
▪Summarize functions and classes
•Generating Onboarding Documentation
▪Draft onboarding guide to get
started quickly
▪Write quick-start examples and
usage instructions
▪Creating documentation for APIs,
functional docs
•Getting Productive Quickly
▪Explain how to run and see
functionality
▪Suggest test cases and
frameworks
▪Reviewing code
•Filling in Gaps
▪Adding docstrings and inline
comments
▪Generating tasks lists
▪Documenting architecture
▪Asking for edge cases for testing
•Managing repo support pieces
▪Summarizing issues, pull requests
▪Suggesting next steps on issues
and pull requests
•Bonus: Quiz me!

20. © 2024 Brent C. Laster &techupskills.com | techskillstransformations.com
© 2025 Brent C. Laster &
@techupskills
21
Prompts used in last slide
Purpose Prompt example
Explain Explain this code in easy to understand terms
Key things to be aware ofWhat are the key files, entry points, and main modules I should be aware of?
Summarize function Summarize the functions and classes in this project
Onboarding Create an onboarding guide for this project
Quickstart/usage Create quickstart examples and usage instructions
User-facing doc Create user-facing functional and API doc
Run and see functionalityExplain how we can best run this code and see the functionality
Testing help Suggest testing frameworks and good test cases - both unit tests and integration tests
Review Review the project code for issues and possible improvements
Task List What are the main outstanding tasks to do for this code?
Architecture with diagramsExplain in simple terms and with diagrams the architecture of the project (mermaidchart.com)
Edge cases What other edge cases should I be testing?
Quiz Create a set of 10 quiz questions that will test my general understanding of this this project and
how it works and then quiz me.

21. © 2024 Brent C. Laster &techupskills.com | techskillstransformations.com
© 2025 Brent C. Laster &
@techupskills
22
|
Testing

22. © 2024 Brent C. Laster &techupskills.com | techskillstransformations.com
© 2025 Brent C. Laster &
@techupskills
23
Testing by direction
•Shortcut
commands/built-in
functions
•Comment-driven
•Prompt-driven

23. © 2024 Brent C. Laster &techupskills.com | techskillstransformations.com
© 2025 Brent C. Laster &
@techupskills
24
Testing by Exploring
•What is the process for
testing and debugging
the project’s code?
•What other edge/test
cases should be
considered?
•How can I test the
performance of X?
•How can I test the
security of X?
•Explain how to create a
test with framework X
for this code?

24. © 2024 Brent C. Laster &techupskills.com | techskillstransformations.com
© 2025 Brent C. Laster &
@techupskills
25
Coverage
•Similar approach applies
to code coverage
•Can ask AI how to get
coverage info
•Provides information
relevant to the language
and needed
dependencies

25. © 2024 Brent C. Laster &techupskills.com | techskillstransformations.com
© 2025 Brent C. Laster &
@techupskills
26
Test-Driven Development

26. © 2024 Brent C. Laster &techupskills.com | techskillstransformations.com
© 2025 Brent C. Laster &
@techupskills
27
Writing integration tests via prompts
•Suggested prompt:
▪ "Write integration tests for the <method/function> in the <class>. Use mocks to
simulate the <external service> and verify that it works correctly when
<condition>."
•Key elements:
▪Scope: Specifies integration tests, focusing on interaction between function and
service, instead of unit tests.
▪Mocks: Explicitly asks for the use of mocks to simulate the service, enabling
interaction with external systems to be tested w/o actual implementation
▪Verification: Prompt emphasizes verifying that service is called correctly based
on condition
▪Specificity: Clearly states the method/function and class to be tested.
Source: https://docs.github.com/en/copilot/using-github-copilot/example-use-cases/writing-tests-with-github-copilot#writing-integration-tests-with-
copilot-chat

27. © 2024 Brent C. Laster &techupskills.com | techskillstransformations.com
© 2025 Brent C. Laster &
@techupskills
28
|
Resolving Issues

28. © 2024 Brent C. Laster &techupskills.com | techskillstransformations.com
© 2025 Brent C. Laster &
@techupskills
29
Resolving issues with AI
•Shortcut fix commands
•What are some places in this repository
where logic/performance/security errors can
occur?
•What are the most recent significant changes
to the codebase?
•What are the code changes that affected this
code in branch x from date y to date z?
•How can I make this code be more efficient
with system resources?
•Where else might this problem occur?
•What other parts of the codebase use this
routine?
•Summarize pull requests, issues, etc.

29. © 2024 Brent C. Laster &techupskills.com | techskillstransformations.com
© 2025 Brent C. Laster &
@techupskills
30
Suggesting next steps
•Given an issue, can
prompt Copilot for
next steps
•May be more
general advice

30. © 2024 Brent C. Laster &techupskills.com | techskillstransformations.com
© 2025 Brent C. Laster &
@techupskills
31
Summarizing issues
•Prompt to
provide summary
of issue
•Includes
overview of issue
based on what
Copilot can
gather from the
system using
tools

31. © 2024 Brent C. Laster &techupskills.com | techskillstransformations.com
© 2025 Brent C. Laster &
@techupskills
32
Leveraging Agents
•Most AI assistants
have agentic
functions built in
•Agent = AI w/
agency
▪ability to
gather input
▪act
▪make changes
•End-to-end
changes
•Autonomous with
oversight
•Goes beyond
typical AI assist
functions

32. © 2024 Brent C. Laster &techupskills.com | techskillstransformations.com
© 2025 Brent C. Laster &
@techupskills
33
|
Documentation

33. © 2024 Brent C. Laster &techupskills.com | techskillstransformations.com
© 2025 Brent C. Laster &
@techupskills
34
Using AI for general documentation via comments
•Basic methods may just do “header” comments
•By being more specific in prompt, can get more extensive comments

34. © 2024 Brent C. Laster &techupskills.com | techskillstransformations.com
© 2025 Brent C. Laster &
@techupskills
35
Generating Framework-compatible documentation
•If the AI recognizes that
the code is in a language
that has a standard
documentation tool it can
create comments ready
for that tool
•Given example code…
•Have AI document it
•Can then generate
framework doc from it

35. © 2024 Brent C. Laster &techupskills.com | techskillstransformations.com
© 2025 Brent C. Laster &
@techupskills
36
Generating Documentation for APIs
•If your codebase is setup for an API
documentation framework
likeSwagger, then you can have the
AI generate the corresponding
Swagger documentation for your
APIs

36. © 2024 Brent C. Laster &techupskills.com | techskillstransformations.com
© 2025 Brent C. Laster &
@techupskills
37
Generating Functional Documentation
•SDAs are also capable
of creating functional
documentation that is
targeted towards
external users
•May also produce a set
of related info including
Identifying Public
APIs,Method
Signatures,Endpoint
Mapping, etc.

37. © 2024 Brent C. Laster &techupskills.com | techskillstransformations.com
© 2025 Brent C. Laster &
@techupskills
38
Extracting Summary Documentation
•SDA can also be used to
extract a high-level
summary of the key
documentation from
certain types of projects
•This can be extremely
useful for quickly
understanding the
significant details of the
code base for a project

38. © 2024 Brent C. Laster &techupskills.com | techskillstransformations.com
© 2025 Brent C. Laster &
@techupskills
39
|
Maintenance

39. © 2024 Brent C. Laster &techupskills.com | techskillstransformations.com
© 2025 Brent C. Laster &
@techupskills
40
Refactoring with AI
•Consider a code example
•Can refactor via prompt
•Resulting code

40. © 2024 Brent C. Laster &techupskills.com | techskillstransformations.com
© 2025 Brent C. Laster &
@techupskills
41
Refactoring for better testability
•Can also ask Copilot to
refactor code to be more
suitable for testing
•For example: " Refactor
the code in
#file:<filename> to make it
more easily testable"

41. © 2024 Brent C. Laster &techupskills.com | techskillstransformations.com
© 2025 Brent C. Laster &
@techupskills
42
Updating with AI
•Can probe with questions
to decide what needs to be
updated?
▪Where do we use API
v1alpha1?
▪What are the changes
between v1alpha1 and
v1 API versions?
•Can prompt to do updates
such as “Change v1alpha1
API to v1”
•AI-recommended change
areas
▪import paths
▪references to API
versions
▪scaffold comments
▪Controller code

42. © 2024 Brent C. Laster &techupskills.com | techskillstransformations.com
© 2025 Brent C. Laster &
@techupskills
43
Augmenting Code with AI
•Augmenting = adding
additional functionality to
existing code
•Prompt “The is_prime
function should log whether
the in put is prime or not”
•AI suggested additional code
that
▪pulls in a logging
framework, initializes a
logger instance
▪adds logging calls for
each possible output

43. © 2024 Brent C. Laster &techupskills.com | techskillstransformations.com
© 2025 Brent C. Laster &
@techupskills
44
Changing Implementations with AI
•Can select code and
prompt for
alternative
implementations
•Consider example
code
•Prompt “Explain this
function in detail
and suggest
alternative
implementations”
•Also - porting
•Prompt “convert to
Go”

44. © 2024 Brent C. Laster &techupskills.com | techskillstransformations.com
© 2025 Brent C. Laster &
@techupskills
45
|
Deployment

45. © 2024 Brent C. Laster &techupskills.com | techskillstransformations.com
© 2025 Brent C. Laster &
@techupskills
46AI assistance in
Deployment
Automation
•Generate CI/CD pipeline scripts
(e.g., GitHub Actions, Jenkinsfiles)
•Assist in writing Dockerfiles,
docker-compose.yml, and
Kubernetes YAML
•Suggest deployment strategies
(blue-green, canary, rolling)
•Detect misconfigurations in infra-
as-code
•Create release notes from commit
history or PRs
•Auto-generate rollback
procedures and health checks

46. © 2024 Brent C. Laster &techupskills.com | techskillstransformations.com
© 2025 Brent C. Laster &
@techupskills
47AI Help with Runtime &
Release Tasks
•Create or refine deployment
documentation
•Recommend versioning
strategies (e.g., SemVer usage)
•Explain output logs and
deployment failures
•Simulate user behavior for post-
deploy validation
•Draft incident response
playbooks
•Help construct monitoring
queries and alerts

47. © 2024 Brent C. Laster &techupskills.com | techskillstransformations.com
© 2025 Brent C. Laster &
@techupskills
48
|
KSummary

48. © 2024 Brent C. Laster &techupskills.com | techskillstransformations.com
© 2025 Brent C. Laster &
@techupskills
49
Key Takeaways & Best Practices
Treat AI like a junior team member
•Always validate and review AI-generated outputs
•Don’t assume correctness; verify like human-written
code
Consider where AI has benefit across the SDLC
•Planning: estimate complexity, prototype ideas, search
indexed knowledge
•Development: generate, refactor, and document code
•Testing: draft unit/integration tests, assess edge cases
•Deployment: write CI/CD scripts, debug configs, generate
release docs
•Maintenance: suggest improvements, spot outdated
APIs, update versions
Choose the right mode of use
•Acceleration: code suggestions inline
•Exploration: ask questions, learn systems
•Directive: issue clear commands for precise
outcomes
Enable with good context
•Index your repos/docs for search (RAG setup)
•Combine embeddings + LLMs for precise answers
Leverage prompt strategies
•Specific, scoped prompts yield better results
•Ask AI to explain, not just generate

49. © 2024 Brent C. Laster &techupskills.com | techskillstransformations.com
© 2025 Brent C. Laster &
@techupskills
50
Pitfalls, Mitigations, and Next-Level
Common Pitfalls
•Blind trust in AI code → introduces bugs or
security issues
•Vague prompts → irrelevant or incomplete
results
•Relying on AI without context → hallucinated
answers
Mitigations
•Implement review gates for AI code
•Prompt for test coverage and verification steps
•Cross-reference AI responses with indexed
documentation
Go Further with AI
•Use AI for summarizing issues, PRs, and
documentation
•Generate playbooks for incident response
•Construct monitoring alerts and validation workflows
•
Enable agent-based tooling (where safe)
•Multi-step, autonomous agents can plan and act
•Best for internal workflows or sandboxed
environments
•
Final Guidance
•Treat AI as a multiplier, not a replacement
•Invest time in prompt design and tool integration
•Balance speed with reliability — quality still matters

50. © 2024 Brent C. Laster &techupskills.com | techskillstransformations.com
© 2025 Brent C. Laster &
@techupskills
51
That’s all - thanks!
techskillstransformations.com
getskillsnow.com
Contact: [email protected]