Case study · SaaS · Generative AI
Improving product gap and revenue analytics with generative AI for PreSkale.
PreSkale · AI-driven presales platform
Identifying product gaps and their revenue impact was manual and slow. We built a Bedrock chain that maps Jira tickets to revenue automatically.
40%
Improvement in product gap creation
25%
Reduction in manual effort
The opportunity
Prioritising features on judgment, from a backlog nobody could read whole.
PreSkale is an AI-driven presales platform that unifies product gap management-identifying and resolving the gaps that block deals. Its product gap analysis depended on someone working through ticket data by hand, which was time-consuming at the volume the platform generates.
Because prioritisation came from interpretation rather than data, mapping a gap to its revenue impact was subjective. PreSkale wanted the analysis itself to be data-driven and to scale past what manual review could cover.
Manual gap identification
Finding critical product gaps meant reading ticket data by hand.
Time-consuming at volume
The process did not scale with the amount of ticket data generated.
Prioritisation was subjective
Feature ranking relied on interpretation rather than evidence.
Revenue impact unmapped
Connecting a gap to the revenue it blocked was done case by case.
The impact
Gaps map to revenue without anyone reading the backlog.
Two chained LLM passes
The first defines ticket context and generates tags; the second classifies ticket groups from them.
25% less manual effort
Tag-based ticket group classification runs automatically rather than being assembled by hand.
Data-driven prioritisation
Features are ranked by potential revenue impact held in the tickets, minimising subjective interpretation.
Insight, not just classification
Classified groups and tags feed the platform's own analytics for customers to act on.
The stack
What it runs on.
- Model layer
- Amazon Bedrock, chained LLM calls
- Source data
- Jira ticket ID, title and description
- Storage
- Amazon S3, CSV format
- Integrations
- CRM, calendar and product
The road ahead
The mapping is automatic. Next is what it can predict.
With gaps classified and tied to revenue as tickets arrive, the same chain extends to anticipating which gaps will block deals rather than reporting the ones that already did.
- 01
Predictive gap prioritisation
- 02
Wider source coverage beyond Jira
- 03
Revenue attribution refinement
Talk to us
Bring one process. Leave with an outcome architecture.
30 minutes on one function, mapped to the architecture behind it. No slides.