Improving Developer Experience using Advanced Platform Engineering Techniques.pdf

This presentation talks about improving developer experience using platform engineering. We first describe the basics of DevEx and PE, and follow that up with value modeling and justifying the investments. We then go into 5 advanced PE Techniques that can used by any organization to provide higher value for your DevEx activities. The presentation concludes by talking about 2 case studies - one from BFSI industry and another from CPG

1. Improving Developer Experience using
Advanced Platform Engineering Techniques


DevOpsDays Cairo 2024
Sep 25, 2024
Ajay Chankramath
Chief Technology Officer & Managing Director
Platforms, Products & Solutions
Brillio, USA

2. Agenda

❖Understand your DevEx goals
❖Basics of applying platform engineering (PE)
❖Establishing metrics needed
❖5 Advanced PE techniques
❖Case Studies

3. Understanding your Developer Experience

Experience for the Developers interacting with Tools,
Frameworks, Process through SDLC
DevEx has all 3 of People, Process & Technology components
which makes it extremely difficult to improve

4. What’s the industry telling us about DevEx?
65% of executives believe improving
DevEx is one of the top goals for
2025
88% of tech executives believe PE is
critical in achieving their software
engineering goals
57% of tech executives believe the
most important thing in PE is
Platform-As-A-Product
58% of digital organizations have
either deployed or deploying a
developer portal by late 2023
Dev Portals + Platform Engineering is
what organizations do.
By 2026, 75% of the organizations
will have that combination, up from
45% in 2024
Various Gartner & Forrester Reports 2024

5. Data Driven DevEx Improvement Model
Qualitative Dev
Assessments
Automatically
instrumented
metrics
Recommended
Practices
VSM/P2P
analysis
Interviews with
engineers
Focus Drivers

Leading &
lagging metrics

PE capability
gaps

Product
capability gaps

Knowledge
management
opportunities

Cultural &
organizational
Roadmap

Org Model

Technical
Options

Change
Management
Build

Paved Roads

Practices

Address Tech
Debt
Gaps/ Opportunity
Identification
Input / Assessment
Techniques
Plan & Strategize Execute

6. Driving action across ecosystem
Onboard Design Build MVP Code Review Deploy


Leading Metrics Lagging Metrics
Qualitative Assessment Quantitative Assessment
Simulate Value to Cost
Map Developer Journey
Connect Metrics
Assess Developer Experience
Value Modeling

7. Platform Engineering



Platform engineering is the practice of designing, building, and
maintaining the underlying ecosystem that enables the development
and delivery of software applications and services.

8. Platform Engineering Vision
PLATFORM ENGINEERING SRE
DevOps
DevEx require supports
enables
Experience for
Developers
interacting with
Tools, Frameworks,
Process through
SDLC
A cultural paradigm that
improves collaboration
and communication
between all aspects of
SDLC
Applying Software
Engineering Principles
to Operations for
creating highly reliable
production systems

9. Platform Engineering Notional View
Developer Plane
Version Control, Infrastructure as Code , Dev Tools, Paved Road

Compliance & Governance Plane
Pipelines , Lightweight governance, FinOps compliance, Compliance @ POC
Delivery & Runtime Plane
Containers, Kubernetes, Workflow orchestration

Networking & Connectivity Plane
VPC, External, 3rd party

Security Plane
IAM, Secret and Encryption Management, SIEM

Platform Product Management
Team Topologies, Technical Product Management,
Value Modeling
Observability
System level, Integrations, Alerting

10. Platform Engineering - Overall Value Proposition

11. How Platform Engineering Helps DevEx?
Source: State of Platform Engineering - 2024 - https://www.puppet.com/resources/state-of-platform-engineering

12. Value Modeling in Platform Engineering
Build what
some pilot
users need
with them
Don’t build
what you can
buy
Show Quick
Value
Tech Product
Management
Proven solutions
keep abstracting
away everything
MVP
Scale only if
you need to
All subsequent
investments
should be aligned
to product life cycle
Value Model

13. Why modern techniques?









As traditional platform engineering techniques become table stakes, more
ideas, also driven by GenAI, need to come to the fore to keep improving value,
differentiating and moving up the abstraction layer.

14. #1 AI Driven Automation
User Research /
Market Research
Requirements
Analysis
Process Analysis
Unit & Integration
Testing
Monitoring &
Observability
Solution Architecture
Coding and
Implementation
Functional & E2E
testing
Quick
WIns
High
ROI
Internal Docs
Regulatory
Guidelines
Vendor APIs
Vendor/Open APIs
Ontologies /
Taxonomies
Research
Publications
Market Data
Market Data
RAGs
Improved
DevEx
Potential Retrieval Augmented Generation (RAGs) to contextualize
your AI models
What you need to do?

15. #2 Multi Cluster Management
Control Plane
Master Node
Data Plane
Worker Nodes
Cloud
Service API
Developers Users
Control Plane
Master Node
Data Plane
Worker Nodes
Cloud
Service API
Developers Users
Control Plane
Master Node
Data Plane
Worker Nodes
Cloud
Service API
Developers Users
Unified Control Plane
Centralized Management, MC orchestration
Automated Cluster Lifecycle Management
Cluster Provisioning / Scaling
Self-Serve Platforms
Workload Orchestration
Optimized Resource Utilization
Multi cluster Networking
Cost Optimization
Quotas / Visibility
Capacity planning
Observability
Cross Cluster Logging
Centralized Monitoring

16. #3 Ephemeral Environments
GitOps
Custom Templates
Pipeline Integration
RBACAuto ScalingObservability
=
Production-Like Ephemeral Envs
Reduce Wastage
On-Demand
Better Quality Products

17. #4 Causal AI Techniques
Observability Platform
proactively detects
patterns
App Instrumented for
Observability
Temporary self-healing
Deep Insights into the
codebase
Code Refactor and generate
code
Real-Time Test Generation
and GatingDeveloper PR

18. #5 Sustainability as a first class citizen
Carbon Tracking Pipeline
based on open APIs
Autoscaling
IaC for Energy Efficiency
Resource Scheduling
Causal AI to rebalance
Architectural Rigor
●Serverless
●Efficient Networking
●Edge Computing
●ML Optimization
●Carbon Offset Tools
Orchestrated Through an IDP

19. Case Study # 1
Problem:
The bank’s SRE functions were heavily focused on
infrastructure, leading to inefficiencies and high
developer overhead. Developers were frequently
called upon to handle L3 incidents, which slowed
down development cycles and impacted overall
product quality and turnaround time.
Solution:
To address this challenge, the bank shifted from a
traditional Platform SRE model to a Product SRE
model, integrating Platform Engineering techniques
to better align SRE with the needs of product
development teams. This approach ensured SREs
became more embedded within the product life cycle,
focusing on system reliability while removing the need
for developers to be involved in L3 issues.
Key changes included:
● Establishing dedicated Product SRE teams for specific business lines.
● Automating infrastructure and environment management using tools
such as Terraform and Ansible.
● Leveraging Sourcegraph Cody for code navigation and ReWrite for
reducing technical debt.
● Using Dynatrace for proactive monitoring and incident detection to
minimize downtime.
Outcomes:
The shift to Product SRE delivered substantial improvements across key
metrics:
1.L3 requirements eliminated: Developers were no longer
required to handle L3 incidents, freeing them to focus solely
on development.
2.Quality improvement: There was an 81% improvement in the
quality of fixes, reducing bugs and system failures.
3.Turnaround time: Incident resolution time improved by an
average of 400%, drastically reducing response times and
accelerating development cycles.
4.Developer Experience (DX): The DX happiness index improved
by a factor of 8X, indicating a dramatic enhancement in the
developer’s overall workflow and satisfaction.

20. Case Study # 2
Problem:
The CPG chain store was facing ESG/regulatory pressure to reduce
its carbon footprint and align with sustainability goals. They set
an ambitious target to cut carbon emissions, but existing
processes lacked the technology and automation required to
achieve this without increasing developer friction or impacting
operational efficiency.
Solution:
The company incorporated sustainability techniques powered by
modern platform engineering technologies to create a more
carbon-conscious development and operational environment. The
key innovation was leveraging automated architectural review
board (ARB) systems that ensured any new architectural decisions
were aligned with carbon-negative or carbon-neutral approaches.
Additionally, automated workflows were introduced to reduce
friction for developers while embedding sustainability as a core
design principle.
Key changes included:
● Automating sustainability assessments in the development pipeline
using CCF and Pachama to monitor and manage carbon impact.
● Pulumi for provisioning and managing infrastructure with automated
sustainability audits.
● Utilizing AWS Lambda and Istio to optimize cloud usage, reduce
waste, and improve resource allocation.
● Cloverly was implemented to help offset carbon emissions through
seamless integrations into operational workflows.
● Backstage was used to centralize and simplify the developer
experience, ensuring that sustainability checks did not add
unnecessary complexity or slow down development cycles.

Outcomes:
1.Carbon footprint reduced by 39%: The organization successfully met
its sustainability goals while continuing to innovate and expand.
2.Automated architecture sustainability assessments: By embedding
carbon-conscious techniques directly into the ARB, decisions about
infrastructure, applications, and deployments automatically aligned
with sustainability targets, without the need for manual intervention.
ARB approvals were faster by 88%
3.Developer friction minimized: Despite the added focus on
sustainability, automation and smart tooling ensured that developers
faced reduced friction, improving efficiency and satisfaction by 21%

21. Takeaways
❖Developer Productivity is an easy topic to complain about and
hard to fix as it involves People | Process | Technology,
precisely in that order

❖Bringing in the rigor of Platform Engineering changes the
equation right off the bat

❖Traditional PE techniques are table stakes, even though there
are lots of organizations still catching up

❖Advanced techniques is what you need for the next 5 years to
stay competitive

22. Stay Connected? Questions?
[email protected]
/chankramath
chankramath.com