This hackathon was managed by Prachi Tiwari, one of 9fin’s directors of engineering - find out more about her role here.
Starting summer on a high, 9finners from across engineering, AI, data science, product and design teams paused their usual work for two days to focus on exchanging ideas and producing practical, valuable solutions to real engineering challenges.
In the spirit of 9fin's core values, a small, lean team of engineers took ownership of the event, schedule, judging criteria, merch (and naturally, the memes to match).
Yes, people had to come to The Office to get merch. But there was no expectation to perform an impromptu dance in reception (we’re so sorry for this joke).
So commenced 9hack, our second company hackathon of 2026. 30 projects were presented across three themes:
- Build the Missing Piece - Ship the feature that unlocks outsized value.
- Win New Markets - Create products that open new revenue opportunities.
- Data to Insight - Turn 9fin data into faster, sharper decisions.
There were no ‘bad ideas’ and no hard constraints on what to build, but all projects had to stick to a few rough principles. We wanted ideas with big impact that addressed a clear pain point. That were built for production, and used AI to move fast. Most importantly, each idea needed a human-in-the-loop who could explain how it works and where it breaks.
Our top two projects: scaling AI company-wide
1. Autocomplete for AI Chat
This smart feature for 9fin's AI Chat that suggests full question completions as users type. It decreases user friction and discovery by reducing typing effort, surfacing questions users might not have considered, and making AI Chat more capable.
Technically, the team used data science to cluster common user questions into normalised "templates" (to protect cross-company privacy), stored them in ElasticSearch for fast retrieval, and wired this to the UI so suggestions appear in real time. The templates include placeholders for company names, which are filled dynamically based on the companies each user follows for personalisation.
Post-hackathon, 9fin’s engineers are actively working on a feature like autocomplete for AI Chat.
2. 9Charts for AI Chat
This feature in AI Chat generates up to 18 types of visual charts using 9fin’s data. Users can see trends and patterns visually rather than interpreting raw numbers in text responses, making data comprehension easier. Users can save charts in the ‘My Charts’ section of their 9fin dashboard.
Technically, the team built a custom LLM "skill" (a rulebook defining which chart types to use and when) paired with a "tool" that takes data and chart specifications and returns a rendered visual. They added a custom markdown renderer plugin to the chat UI so charts appear inline in the stream, with a toggle back to the underlying data table. The charting library is Apache ECharts, which also enables PNG, HTML, and CSV export.
9Charts is included in 9fin’s 2026 engineering roadmap, so look out for this feature soon!
Special mention: Sales Lead Generation Tool
This custom internal tool uses 9fin's articles to enhance sales efficiency and discovery by helping the sales team find new prospective clients, identify existing accounts worth re-engaging or upselling, and arm sales representatives with relevant context for outreach.
The tool surfaces firms, desks, sponsors, and deal data mentioned across 9fin's articles and holdings, enriches lead rationale with Salesforce interaction history and internal content, and provides filtering by assignment status, region, and team type.
This tool could be included in the 2026 roadmap, pending further tinkering.
What's in store for our next hackathon?
Hackathons give us a chance to explore new ideas quickly. Some of these projects will make their way onto our roadmap, and others may inspire entirely new directions. Either way, 9hack proved that some of the best ideas emerge when talented people are given the freedom to explore them.
We’ll soon be announcing the themes for our next hackathon — stay tuned!
9fin’s engineers are solving hard problems at the intersection of finance and AI. If that sounds like your kind of challenge, come build with us.
