Gen Digital says its Norton Revamp platform uses Amazon Bedrock AgentCore Web Search to ground personalised news feeds and professional-content suggestions in current web results. The practical idea is to give advice about what to post a live view of what is happening, rather than relying on a static store of yesterday’s knowledge.
AWS AI Watch analysis
What happened
Norton Revamp launched in March 2026, according to a case study published by AWS. When a user sets a professional goal, the system generates targeted search queries, retrieves results with source URLs and publication dates, then ranks them for relevance, recency and source credibility. Those results inform news feeds, content suggestions and weekly goal-setting.
The case study describes a four-stage workflow: an initial scan of up to 1,000 historical posts, daily updates, weekly goal generation and a coaching flow that moves from analysing an idea to refining and posting it. Gen Digital says the product has had 99.9% uptime since launch. It also reports that AI helped make feature implementation three times faster, while review of AI output still took 20–30% of the team’s time.
Why it matters
This is a concrete example of search being used as a live information layer for a generative AI product, not just as a chatbot feature. The design also makes clear that retrieval alone is not the whole trick: the team says it ranks sources and feeds user engagement back into future searches.
The figures come from Gen Digital’s account of its own product, published by AWS, so they are company-reported results rather than an independent performance assessment. Even so, the workflow gives builders something more useful than a slogan: a description of where search, ranking, generation and human review fit.
Our read
The promising bit is the connection between current sources and a specific task. A feed that can show its sources has a better shot at being useful than a confident content machine working from stale information. The less glamorous lesson is that someone still has to judge the results and review the output. The automation has not abolished editorial judgement; it has given it more tabs.
What to watch
- Whether Gen Digital publishes further detail on how it measures relevance and source credibility.
- Whether the reported user engagement feedback improves the quality of recommendations over time.
- How much human review remains necessary as the system scales.
Discussion spark: For AI-generated professional advice, is showing cited current sources enough to earn trust, or should the system also explain how it chose and ranked them?
Sources and evidence
- How Gen Digital built Norton Revamp’s AI-curated News Engine with Amazon Bedrock AgentCore Web Search (3 November 2026, 16:43 UTC)
- Introducing India-based inference for Kimi K3 on Amazon Bedrock (3 November 2026, 16:43 UTC)
- Artificial Intelligence (3 November 2026, 16:43 UTC)
not affiliated with or endorsed by Amazon Web Services (AWS)