AI Daily Signal
Built an internal intelligence product that identifies and summarises important AI developments for senior decision-makers.
Problem
Senior leadership needed to stay on top of fast-moving AI developments relevant to the business, but had no reliable, low-effort way to do it beyond scattered newsletters and conference talks.
What I built
A daily digest pipeline that scrapes, filters, scores, and summarises AI news and research into a ranked brief delivered straight to Slack every morning.
My contribution
Designed and built the entire pipeline solo — source selection, relevance scoring, summarisation prompts, and the Slack delivery format — then iterated weekly based on direct leadership feedback.
Process
Mapped the real behaviour
Shadowed how leadership actually tried (and failed) to keep up with AI news today.
Built a scoring model
Weighted sources and topics so the signal-to-noise ratio stayed high as volume scaled.
Tuned summarisation to the reader
Prompt-engineered a summarisation layer calibrated to an executive reading level, not a researcher's.
Shipped and iterated weekly
Launched a daily Slack digest and adjusted format and tone based on direct reactions.
Outcome
Adopted by the leadership team as their primary way of tracking AI developments; cut weekly research time from roughly 4 hours to 20 minutes.