Ep080: When AI Meets Accounting - How Sage is Transforming Business Software

02/25/25 • 20 min

From cost management to practical implementation, Sage's Amaya Souarez shares invaluable insights on building AI-powered business tools that deliver measurable value to customers.

Topics Include:

Amaya Souarez introduced as EVP Cloud Services at Sage
Overview of Sage: offers accounting, finance, HR and payroll tech for small businesses
Company emphasizes human values alongside technology development
Amaya oversees core cloud services and operations across 200+ products
Sage Co-Pilot announced as new AI assistant – helping automate invoicing and cash flow management
Common misconceptions with Generative AI
AI solutions aren’t always solution to every problem
Compares AI hype to previous blockchain enthusiasm
Emphasizes starting with clear use cases before implementation
Difference between task-based and reporting-based use cases
Partnering with AWS to build accounting-specific language models
Different accounting terminology varies by country
Using AWS Bedrock and Lex for a domain-specific language model development
Multiple AI models may be needed for single solution
Customer feedback drives project funding decisions
AI development integrated into regular product roadmaps
Focus on reducing cost per user for AI features
Success story: reducing 20-hour task to 5 minutes
Tracks AI usage costs per customer interaction
Early Gen AI hype caused confusion in the market
Plans to make domain-specific models available via API
Will offer language models on AWS Marketplace
Emphasizes practical AI application over blind implementation

Participants:

Amaya Souarez - EVP Cloud Services and Operations, Sage

See how Amazon Web Services gives you the freedom to migrate, innovate, and scale your software company at https://aws.amazon/isv/

From cost management to practical implementation, Sage's Amaya Souarez shares invaluable insights on building AI-powered business tools that deliver measurable value to customers.

Topics Include:

Amaya Souarez introduced as EVP Cloud Services at Sage
Overview of Sage: offers accounting, finance, HR and payroll tech for small businesses
Company emphasizes human values alongside technology development
Amaya oversees core cloud services and operations across 200+ products
Sage Co-Pilot announced as new AI assistant – helping automate invoicing and cash flow management
Common misconceptions with Generative AI
AI solutions aren’t always solution to every problem
Compares AI hype to previous blockchain enthusiasm
Emphasizes starting with clear use cases before implementation
Difference between task-based and reporting-based use cases
Partnering with AWS to build accounting-specific language models
Different accounting terminology varies by country
Using AWS Bedrock and Lex for a domain-specific language model development
Multiple AI models may be needed for single solution
Customer feedback drives project funding decisions
AI development integrated into regular product roadmaps
Focus on reducing cost per user for AI features
Success story: reducing 20-hour task to 5 minutes
Tracks AI usage costs per customer interaction
Early Gen AI hype caused confusion in the market
Plans to make domain-specific models available via API
Will offer language models on AWS Marketplace
Emphasizes practical AI application over blind implementation

Participants:

Amaya Souarez - EVP Cloud Services and Operations, Sage

See how Amazon Web Services gives you the freedom to migrate, innovate, and scale your software company at https://aws.amazon/isv/

Previous Episode

Ep079: AI, Innovation, and the Enterprise: Box's Journey with AWS

Box's Chief Product Officer Diego Dugatkin discusses how the enterprise content management platform is leveraging AI through partnerships with AWS Bedrock and continuing to innovate for their customers.

Topics Include:

Introduction of Diego Dugatkin as Box's Chief Product Officer
Box provides cloud content management for enterprise customers
Focus on Intelligent Content Management
Box serves 115,000 customers including 70% of Fortune 500
Company manages approximately one exabyte of enterprise data
Box expanding product portfolio to offer more customer value
Partnership with AWS Bedrock for AI implementation announced
Collaboration with Anthropic for LLM technology integration
Box offers neutral approach letting customers choose preferred LLMs
Common misconceptions about generative AI capabilities and limitations
Generative AI helps accelerate contract analysis and classification processes
Box Hubs enables content curation and multi-document queries
Success measured through hub creation and query accuracy metrics
Long-term AWS partnership continues expanding with new technologies
Amazon is major Box customer while Box uses AWS
API integration important for third-party developer implementations
AI development exceeding speed expectations in efficiency improvements
Challenges remain in defining AI agent roles and capabilities
Content strategy crucial for deploying intelligent content management
Companies must prepare for AI agents in workplace
Flexibility in tech stack recommended over single-vendor approach
Next 12-24 months will see accelerated industry changes
Box maintains innovative culture through intrapreneurship approach
Company regularly hosts internal and external hackathons
Focus on maintaining integrated platform while acquiring companies
Partnership between Box and AWS continues growing stronger

Participants:

Diego Dugatkin – Chief Product Officer, Box

See how Amazon Web Services gives you the freedom to migrate, innovate, and scale your software company at https://aws.amazon/isv/

Next Episode

Ep081: Customer-First AI: DTEX Systems’ Journey with Generative AI and AWS

Ryan Steeb shares DTEX Systems’ strategic approach to implementing generative AI with AWS Bedrock, reducing risk while focusing on meaningful customer outcomes.

Topics Include:

Introduction of Ryan Steeb, Head of Product at DTEX Systems
Explanation of insider risk challenges
Three categories of insider risk (malicious, negligent, compromised)
How DTEX Systems is using generative AI
Collection of proprietary data to map human behavior on networks
Three key areas leveraging Gen AI: customer value, services acceleration, operations
How partnership with AWS has impacted DTEX's AI capabilities
Value of AWS expertise for discovering AI possibilities
AWS Bedrock providing flexibility in AI implementation
Collaboration on unique applications beyond conventional chat assistants
AWS OpenSearch as a foundational component
Creating invisible AI workflows that simplify user experiences
The path to monetization for generative AI
Three approaches: direct pricing, service efficiency, operational improvements
Second and third-order effects (retention, NPS, reduced churn)
How DTEX prioritizes Gen AI projects
Starting with customer problems vs. finding problems for AI solutions
Business impact prioritization framework
Technical capability considerations
Benefits of moving AI solutions to AWS Bedrock
Fostering a culture of experimentation and innovation
Adopting Amazon's "working backwards" philosophy
Balancing customer-driven evolution with original innovation
Time machine advice: start experimenting with Gen AI earlier
Importance of leveraging peer groups and experts
Future outlook: concerns about innovation outpacing risk mitigation
Security implications of Gen AI adoption
Participation in the OpenSearch Linux Foundation initiative
Final thoughts on the DTEX-AWS partnership