Tempo: A monetisable Digital Asset Management system
Project name: Digital Asset Management (DAM) and Tempo Assistant
Newsroom: Tempo (Indonesia)
Solution: A structured DAM system organising 55 years of archives for commercial licensing and powering a reader chatbot
Organisation staff size: 21 - 50 staff
Tech stack: Custom DAM architecture, retrieval-grounded LLM chatbot, automated media data feeds
Impact: Transformed archives into a commercial data monetisation pipeline and provided reader synthesis with zero editorial friction
Core lesson: Fixing metadata and data architecture is 80 per cent of the work; design AI workflows to add zero friction for reporters
Tempo is an independent media organisation based in Indonesia with a storied history. It has over 50 years of archives in investigative journalism, rare photos, and reporting that provide deep historical context. On the newsroom end, they realised they had inherited a library where half the book jackets were missing, a few photos just said “Jakarta, late 90s”, and rights were scattered across different formats. The problem was compounded for audiences.
The problem: A disorganised archive
“On the audience side, we had a double-barreled problem: attention span and trust. Modern readers don't want to type keywords into a clunky search bar and sift through 50 blue links from 2012 to understand a complex political backstory. They want instant synthesis. But, at the same time, audiences are deeply sceptical of AI-generated fluff,” said Gregorio Vincent Wijaya, Senior Product Manager at Tempo.
If they couldn't deliver fast, conversational answers anchored in verified facts, they risked losing the next generation of readers to generic social media summaries, Wijaya added.
Building the solution: A digital asset management and monetisation tool
Before they could go further, they began with what Wijaya called a “massive metadata intervention”.
“If you push chaotic data into a modern platform, you don't get innovation, you just automate the chaos. Second is people: there are lots of iterations to fulfil every need from every team within the organisations. Third, is tools: we couldn't afford an AI that acts like an overly confident intern making up facts when it doesn't know the answer,” explained Wijaya.
To solve this, they engineered a Digital Asset Management (DAM) system with a strict, retrieval-grounded architecture. They put the assistant on a strict news-only diet: it synthesises answers strictly from verified Tempo articles. If the answer wasn't sitting right there in their archives, the bot simply admits it doesn't know, rather than validating or impressing with a hallucination.
Their primary guiding decision was "zero additional friction" for the newsroom. Journalists operate under strict deadlines; if a new tool adds even three extra clicks to their publishing routine, adoption will fail. Therefore, they decided to fully decouple the AI and data monetisation workflows from the editorial publishing workflow.
The DAM is connected to their data platform and aids Tempo in multiple ways. It sells their data so that AI and other parties that want the data can acquire them through a commercial agreement, Wijaya said. For audiences, the result is the chatbot Tempo Assistant, which helps readers understand and engage with their content with ease. The DAM also reduced hassles. For example, the visuals team no longer needed to provide manual approvals for photo use by the rest of the organisation.
“For the data platform and chatbot, we are creating an automatic system that enables any integrated media data feeds into the data platform when the database is updated (meaning any designated articles or content that will be able to be purchased are published). And the data platform is used as a source of knowledge for our chatbot. So basically there are no additional processes required from the newsroom,” Wijaya added.
The opportunities: Focused AI strategy and implementation
The DAM tool development process helped them shift from seeing AI research and development as a luxury they couldn’t afford to seeing it as an executable strategy.
“As an independent media organisation, our margins are tight, and resources are fiercely guarded for core journalistic operations. The JournalismAI Innovation Challenge grant provided the financial runway to dedicate engineering hours and secure the initial infrastructure without cannibalising our newsroom budget,” said Wijaya.
It allowed them to transition from a defensive posture to an offensive one, he added. They no longer worried about not being compensated when AI companies accessed their historic archives and instead built a data monetisation platform.
Furthermore, the prototype they built with this grant served as the critical POC needed to secure their subsequent partnerships, like the GPU credits for local LLMs and the co-development agreement with their Malaysian partners. It essentially catalysed Tempo’s shift from a traditional media company to a media-tech organisation.
Lessons for newsrooms
AI is only as good as the underlying metadata: You cannot simply plug an LLM into a messy, 55-year-old archive and expect magic; fixing the data architecture and enforcing strict tagging protocols was 80 per cent of the actual work.
Change management is harder than software engineering: They had to prove to their photo desk and journalists that the DAM would save them time in the long run to overcome the initial resistance towards new upload protocols.
In journalism, an AI tool saying "I don't know" is a feature, not a bug: By heavily constraining its chatbot to only answer based on verified Tempo archives, the team realised that transparency builds immense trust with their audience. “Readers appreciate a tool that refuses to guess, reinforcing our brand as a reliable, ethical source of truth in an era of generated content, said Wijaya.
Explore case studies from the 2025 cohort
The 2025 Innovation Challenge Report, Prompting Sustainability, collects case studies from all 12 grantees and reveals their journeys, challenges, and opportunities.
Read about their experiences, discover best practices, and find inspiration in their innovative approaches to integrating AI into journalism.
The JournalismAI Innovation Challenge, supported by the Google News Initiative, is organised by the JournalismAI team at Polis – the journalism think-tank at the London School of Economics and Political Science, and it is powered by the Google News Initiative.
