How to Make a Presales Cloud Architecture Proposal

Angelo Muñoz
12 min readNov 15, 2022

“Approach each customer with the idea of helping them solve a problem or achieve a goal, not of selling a product or service.” — Brian Tracy

Photo by Gabriel Crismariu on Unsplash

Presales solution architects are an important part of any cloud service sales process. The presales architect is usually tasked with producing a technical proposal that will satisfy a potential customer’s business needs. As such, a great presales architect will have good listening and communication skills to properly understand, interpret, and document the business requirements.

With the proper understanding of the customer’s challenges and opportunities, the presales architect then goes to work preparing a technical solution and a proposal.

A good proposal will have an executive summary, requirements, proposed scope of services, proposed solution, your company’s profile, delivery model, and cost. These important sections establish customer trust by:

  • Summarizing the challenges and opportunities in the executive summary
  • Detailing the deliverable with requirements and proposed scope
  • Explaining how your solution meets the business need with the proposed solution
  • Demonstrating how you will deliver the solution in the delivery model
  • Showing the one-time implementation cost, as well as the ongoing operational cost of your proposed solution
  • Summarizing the challenges, opportunities, and the proposed solution in your conclusion

To demonstrate these points, I share a sample cloud architecture proposal using a fictional customer, GEM Smart Technologies, and fictional cloud architecture firm, AMC. It assumes the customer has a corporate policy of using AWS serverless technologies.

Following is the proposal:

GEM Smart Technologies

GEM Electron Serverless Backend Migration
Prepared by: Angelo Muñoz

Executive Summary

GEM Smart Technologies (Customer) is a provider of consumer IoT products. Their most recent innovation, the GEM Electron, is a water-powered device that produces enough electricity to power a 400 square foot space (or 20x20 ft room).

Homeowners monitor their GEM Electrons using a mobile app, which shows electricity generated, cost savings over local utility companies, usage patterns, and uptime. Since its launch 1 year ago, GEM Electron sales have doubled every month, which has overwhelmed the backend systems, leading to several outages. This led to lost customer trust, 10% drop in sales, and brand damage as customers complained heavily on online forums and social media.

To remedy this challenge, GEM has partnered with AMC, a cloud architecture firm, to design and implement a cloud-based backend solution to allow GEM to regain customer trust, maintain its market lead, and satisfy its growing customer base.

AMC will work with GEM to understand their business goals, key performance indicators, industry regulatory or compliance needs, business continuity requirements, and technical requirements. AMC will design and implement a new serverless backend to the agreed-upon specifications and work with GEM business and technical contacts to test and refine the solution. Included in the scope will be a zero-downtime migration strategy from the existing infrastructure to the new, as well as tooling to enable rapid application enhancements for business agility. The existing IoT backed solution will remain as is.

Throughout the project, AMC will also engage with the GEM technical team for knowledge transfer.

Using AMC’s proposed solution, similar customers have cut costs by an average of 60% while realizing a 30% revenue increase.

Requirements

Business Requirements
The following business requirements were identified during the discovery phase:

  • Return device status reliably
  • 99.9% availability
  • Respond to customer requests as fast as possible while considering costs
  • Maintain regulatory compliance
  • Decreased infrastructure costs
  • Support all US time zones
  • Centralized operational visibility

Key Performance Indicators
The following Key Performance Indicators were identified during the discovery phase:

  • Monthly Customer Cost Savings
  • Devices in Operation
  • Customer Satisfaction (out of scope of this proposal but listed for reference as a potential future project)

Technical Requirements
The following technical requirements were identified during the discovery phase:

  • Provide consistent logging, tracing, metrics and alerting for observability
  • Meet the 2 second max latency API response
  • API to support 30 requests per second, for up to 1 KB payload size
  • Sensors report status every 30 seconds (informational to support cache TTL config)

Proposed Scope of Services

AMC will provide the following Implementation Services:

I. Activities

AMC Project Process

a. Plan
i. Identify product owner, key stakeholders
ii. Translate business requirements into technical requirements (backlog)
iii. Agree on collaboration and communication strategy
iv. Address risk

b. Learn
i. Review customer goals and challenges
1. Review business requirements
2. Review/define successful outcomes
3. Review/define technical requirements
4. Review Key performance indicators (KPIs)

ii. Review existing backend implementation
1. AWS Account strategy
2. Review Technical challenges
3. Application Architecture
4. Network, security architecture
5. Operations monitoring
6. Review Data Strategy

iii. Review Business Continuity, Disaster Recovery needs
1. Review Data classification, retention
2. Review Recovery Time Objective, Recovery Point Objectives

iv. Review Deployment needs
1. Infrastructure as Code (IaC) requirements
2. Current state of processes and tools
3. Review CI/CD requirements
4. Review testing requirements
5. Environment promotion process

c. Design

i. Authentication/Authorization using Cognito
1. Open ID Connect (OIDC), Cognito integration
2. Define Authorization Scopes

ii. REST API
1. Methods: Get Device status, register device
2. Decoupled architecture: Use SQS for Queues

iii. Availability/Reliability
1. Use multiple availability zones for Lambda functions
2. Use services with high-availability built-in (API Gateway, DynamoDB, WAF, Route53, SQS)
3. Guaranteed delivery with exponential backoff using jitter
4. Automated pre-release integration testing in pipeline

iv. Caching for performance
1. API Gateway caching
2. TTL

v. Network
1. Design VPC IP scheme for architecture expansion and autoscaling headroom
2. NACL’s with least-privilege access
3. Security Groups with least-privilege access
4. Internet Gateways

vi. Database
1. Use DynamoDB single table design
2. Define access patterns
3. Design Data model (Partition, Sort keys)
4. Design Indexes (LSI, GSI)

vii. Compute (Serverless)
1. Lambda
2. Velocity Template Language (VTL) to save compute costs

viii. Security
1. Firewall rules
2. WAF
3. Encryption
4. Access control

ix. Automated deployment pipelines
1. Design build/deployment model in partnership with Customer
2. Design release pipeline templates for future reuse
3. Design rollback/roll forward strategy
4. Design feature-flags processes using Launch Darkly

x. Business Continuity/Disaster Recovery
1. Solution to meet RTO 5 mins, RPO 45 mins
2. Use DynamoDB global tables for cross-region data replication

xi. Ongoing enhancements
1. Feature flags
2. Automated testing

xii. Operations
1. Observability — CloudWatch/CloudTrail
2. KPI Dashboards

xiii. Cutover
1. Blue Green deployment via DNS (weighted routing gradual cutover after validation tests)
2. Automated testing promoted through low to higher environments

d. Implement
i. Build Infrastructure
ii. Build business logic
iii. Build CI-CD pipelines with Microsoft Azure DevOps for automated pipeline deployment
iv. Build Cutover plan

II. Deliverables: AMC will work to complete these deliverables (“Deliverables”):
a. AWS Serverless backend
b. Zero-downtime backend migration from current to new backend
c. Microsoft Azure DevOps Deployment pipelines
d. Solution Runbooks and other implementation documentation developed throughout the project
e. Architecture diagrams and other diagrams created throughout the project
f. All templates, scripts, source code, and other artifacts developed throughout the project
g. Out of scope: IoT backend solution

III. Components: The Services and Deliverables will be limited to the following components:

All components are in AWS unless otherwise stated.
a. Route53
b. WAF
c. API Gateway
d. Cognito
e. Lambda
f. SQS
g. DynamoDB
h. AppSync
i. S3
j. CloudWatch
k. CloudTrail
l. Guard Duty
m. Config
n. Quicksight
o. Microsoft Azure DevOps for deployment

IV. Assumptions: In performing the Services, AMC is making the following assumptions:
a. The backlog prepared and agreed-upon jointly by Customer and AMC will be the scope of this project
b. Access and permissions to the necessary AWS accounts and/or internal systems will be granted during project kickoff or before
c. IoT backend service is complete and functioning to source data to DynamoDB
d. Documentation stored in repo for easy customer access
e. Limited to mobile app backend; IoT backend out of scope

V. Design Considerations: AMC and Customer agree to the following:
a. Solution architecture should use AWS managed services, where possible, to follow Customer’s corporate standard
b. Solution architecture and implementation should be secure, scalable, highly available, and cost-effective
c. Security should be included in every phase of the solution
d. Solution will include automated deployment tools for both software build and release pipelines, as well as infrastructure using Infrastructure and Code (IaC) to ensure a repeatable and rapid release process
e. Architecture and implementation decisions will be influenced by the AWS Well Architected Framework

VI. Customer Responsibilities: AMC is relying on Customer for:
a. Making available key stakeholders, decision makers, operations personnel, development staff, subject matter experts, security staff, and other business and technical staff in a timely manner
b. Appropriate access to the necessary AWS accounts to perform the activities and implement the deliverables involved in this engagement
c. Timely onboarding onto Customer systems as needed
e. Obtaining necessary 3rd party application licenses

Proposed Solution

GEM Technology seeks to regain customer trust, maintain its market lead, and satisfy its growing customer base. To meet these needs, AMC proposes the following solution.

Serverless backend

Following GEM’s policy of using serverless technologies, AMC proposes a new serverless backend using AWS API Gateway, Lambda, and DynamoDB.

This solution will meet the requirements in the following ways:

  • Return device status reliably. With service-level agreements, API Gateway, Lambda, and DynamoDB provide uptime guarantees, or AWS will refund infrastructure costs. In addition, AMC will use automated deployment processes to quickly remediate issues caused by deployments. In addition, AMC will put observability monitoring in place which will notify the GEM operations team of potential issues before they turn into outages. And as a final layer, AMC will implement uptime monitoring that will alert GEM’s operations team of site outages or issues before customers report them.
  • 99.9% availability. Per AWS Architecture recommendations, AMC will deploy the backend across multiple availability zones, and for added measure will deploy across 2 regions to preclude issues from a single region outage.
  • Respond to customer requests as fast as possible while considering costs. With API Gateway caching, the solution will place data closer to the customer, reducing latency and improving the customer experience. AMC estimates average response times improving by as much as 60% over the on-prem solution.
  • Maintain regulatory compliance. The in-scope AWS Services are ISO-27001 certified.
  • Decreased infrastructure costs. With pay-as-you-go pricing, free tier, and free data ingestion, the solution is estimated to cut infrastructure costs by over 70% compared to the current on-prem solution.
  • Support all US time zones. AMC will deploy the solution in Oregon and Virginia in the US to provide increased performance and for disaster recovery purposes. The API business logic will also incorporate time zone logic supporting all US time zones.
  • Centralized operational visibility. With AWS CloudWatch, GEM’s operations group will have a centralized operations dashboard showing key metrics such as API error rates, network throughput, and database latency, among others.

Architecture

Below are 2 architecture diagrams with the proposed solution. The first is a high-level diagram, while the second is a detailed diagram.

Device owners use their mobile app which connects to a serverless backend using API Gateway with an application layer (AWS Lambda) and a database (DynamoDB), which receives device updates from AWS IoT Core.

For product analytics, DynamoDB streams results to an S3 bucket, which serves as a data source to Amazon Quicksight, a business intelligence and analytics tool.

GEM Electron Serverless Backend Architecture Diagram

The detailed solution shows a multi-region, active-active architecture using Amazon DynamoDB global tables.

GEM Electron Serverless Backend Architecture Diagram — Detailed

AMC Profile

AMC is a leading, global, cloud-native services provider. We’re committed to helping our customers adopt, thrive, and succeed in the cloud. As an AWS Premier Consulting Partner, we deliver industry-leading service by our more than 300 consultants with over 1,500 AWS certifications with an average of 8 years’ experience delivering successful solutions in the cloud.

We bring capabilities such as cloud migration, cost optimization, cloud-native development, application modernization, Internet of Things (IoT), and machine learning.

With a successful delivery record across industries such as manufacturing, energy, healthcare, automotive, travel, and other industries, we can help you accelerate getting value from the cloud.

Customer Success

AMC’s customers are from a wide range of industries, including healthcare, manufacturing, transportation, communication, and energy among others. They range in size from startups to Fortune 500 enterprises.

AMC Serverless Prior Expertise

AMC is pleased to provide the following customer references of similar scope and complexity to (insert reference project name here).

See the case study (insert link to case study) for more information.

Delivery Method

AMC utilizes an agile delivery framework which provides ongoing stakeholder engagement and visibility. This approach uses a 1-week feedback cycle which fosters continuous feedback and corrections throughout the project.

Review

Each 1-week cycle is known as a sprint. Before each sprint, AMC will collaborate with Customer in a planning session to prepare “user stories”, or deliverable portions of a product or project. During this collaboration session, typically known as a “backlog review”, Customer will respond to and clarify any outstanding unknowns. Customer will also prioritize the stories, with the highest priority stories going to the top of the backlog. This priority allows AMC engineers and architects to select Customer’s highest priority work to work on during the sprint.

Planning

On the first day of the sprint, AMC and Customer will meet to review the user stories that AMC architects and engineers can deliver in the coming week. This session is called “sprint planning”.

During the Sprint

During the sprint, AMC engineers will deliver the analysis, design, implementation, and testing for the stories in the sprint. They will also meet daily for a brief “standup” meeting to discuss what each team member accomplished yesterday, what they will accomplish today, and any roadblocks in the way.

Demonstration

At the end of each sprint, AMC will hold a demo showing the work completed during the sprint. This demo allows stakeholders to view progress and provide feedback.

Continuous Improvement

AMC will also discuss at the end of each sprint what went well, and what can be improved for next sprint. This internal feedback loop allows the team to make frequent, small, and doable adjustments that impact the project in a positive way.

Customer Acceptance

AMC is responsible to deliver a successful solution for Customer. At the end of the sprint demo, Customer will declare if the work delivered meets the stories’ requirements. Stories deemed satisfactory will be marked complete. Stories deemed incomplete will have the remaining work carried over into the next sprint.

If Customer identifies new work, new stories are created and reviewed in the backlog review meeting.

Resource Profiles

AMC will assign the following roles to the project.

Managing Consultant
Is responsible for overall project success. Leads all aspects and phases of the project, including plan, learn, design, and implement, finances, and contracts. In addition, this professional owns the strategic and long-term planning with Customer.

Solutions Architect
Is responsible for the successful delivery of the technical solution. This professional converts the business and technical requirements into an architecture and leads the implementation teams to build that solution. They partner very closely with the implementation team to resolve any technical challenges, mitigate risk, and ultimately deliver a successful outcome.

Cloud Engineer
Is responsible for implementing the technical architecture with software, infrastructure as code, and continuous integration/continuous deployment pipelines. This professional delivers quality software by writing reliable, automated tests following test-driven and behavior-driven development, starting with the end in mind and working back from that point to the application logic.

Project Manager
Is responsible for ensuring the efficient operation of the project. This professional oversees project planning and execution, and removes roadblocks, and facilitates sprint meetings.

Estimated Pricing

AMC will invoice for services using a “time and material” model, where Customer will pay for time spent and any licenses or materials.

Amazon Web Services (AWS) has provisionally committed to partially fund this project up to the amount listed below.

Based on the Customer’s requirements, AMC is providing the estimate below.

One-time Cost:

  • Estimated Project Time: 10 weeks
  • Estimated Cost: $150,000
  • AWS Discount: $15,000
  • Net Total: $135,000

Operational Cost:

  • Estimated AWS Operational Cost*: $400/mo

* This cost was estimated using current rates (as of the date of this estimate) and is subject to change by AWS; actual costs will vary.

Conclusion

AMC is a proven AWS Premier Consulting Partner with over 50 customers in the US, Europe, and Asia. With a wealth and depth of experience that AMC brings, you will feel confident in AMC’s ability to deliver a quality solution that meets the needs of your business to restore customer confidence and continue positive growth. AMC looks forward to the opportunity to partner with GEM Technologies to build a reliable, secure, scalable backend to meet its business objectives.

--

--

Angelo Muñoz

Enterprise Cloud Architect with 20+ years' experience delivering business and technology solutions for various industries. #customer-first #leadership #business