Aura Health: turning AI into a working marketing engine.
We implemented Claude workflows and a multi agent creative engine for Aura’s growth and marketing team. The engine connects competitor research, Meta ad generation, brand review and campaign performance feedback, giving the team a repeatable route from brief to creative testing.
- Client
- Aura Health
- Sector
- Digital wellbeing & growth operations
- What we built
- A multi agent marketing engine and Claude workflows
- Headline outcome
- Research, creative production and review connected
More campaign ideas still meant more production work.
Growth teams need to study the market, develop a brief, produce creative variations and check whether each one fits the brand. Those steps can consume the time needed to learn from the next campaign.
Aura needed useful AI implementation around that recurring work. The engineering task was to connect research, generation, review and performance signals into a repeatable process, rather than leave the team moving outputs between separate tools.
Research, generate, review and learn.
Competitor research as a workflow input
Research agents monitor competitor campaigns and surface messaging patterns and market gaps for the creative process.
A generative creative engine
The system produces targeted static and video ad variations for Meta campaigns, taking the work from a brief into creative options the team can review.
An AI critic in the production flow
A critic agent checks generated ads against brand guidance and performance practices. Review is a stage in the workflow, rather than a separate task after generation.
Performance feedback and team enablement
Campaign performance data feeds the scaling loop. Alongside the engine, Claude implementation supports the growth and marketing team’s recurring AI work.
The output of one stage becomes the context for the next.
Defined agent responsibilities
Research, creative generation and critique have distinct roles inside the marketing workflow.
Brand context in the review
Creative checks use the team’s brand guidance so generation and evaluation share the same frame of reference.
Feedback beyond generation
The engine connects creative production to live performance signals, allowing the workflow to learn from what runs.
Built around the marketing lifecycle.
The implementation follows the work from market research through creative production and review to performance feedback. Claude sits within the team’s broader enablement work.
- 01
Connect research and context
Bring competitor signals and the creative brief into the production process.
- 02
Build generation and critique
Create variations and evaluate them against the relevant brand and campaign guidance.
- 03
Close the feedback loop
Use campaign performance information to inform the next production and scaling decisions.
A repeatable creative workflow inside the team.
research, generation, critique and performance feedback
for producing and testing brand aligned variations
Claude implementation around growth and marketing work
A few useful
questions.
Was this only prompt training?
The engagement included implementation work: a multi agent creative engine alongside Claude use for the growth and marketing team.
Are these results from Aura’s wellbeing app?
This case covers the marketing and growth workflow. Aura’s consumer platform and voice companion work are separate engagements.
A custom property platform connecting onboarding, tender selection, follow up and payments.
Read the case study
Building an owned, bilingual platform for scheduling, service reports, invoicing and customer visibility.
Read the case studyWhat would change if AI could carry the work from brief to review?
From the first proof to a working implementation. We’ll explore the workflow, the systems involved and a practical first build.