Why We Backed Matterfact at z21 Ventures
Every investment hides behind a mountain of noise.
To form a meaningful thesis on a mid-cap stock, analysts often wade through 10,000+ pages of SEC filings, earnings transcripts, industry reports, and expert-call notes. This research cycle can stretch 30–40 days — and that’s before a single dollar is deployed.
Post-investment, the “sacred cow” syndrome kicks in. Teams spend 2–3 hours a day per position tracking incremental updates — sector dynamics, peer benchmarks, channel checks — just to stay current.
This manual effort means one thing: the opportunity cost of research is rising.
Analysts spend more time maintaining coverage than finding new ideas or challenging old ones.
The Alpha Bottleneck
Despite investment teams pouring in hundreds of analyst-hours every quarter, coverage of mid-cap companies is shrinking. Wall Street’s research budgets have been slashed. Many buy-side firms lack the capital or appetite to build internal data science teams.
This creates a systemic constraint: an information bottleneck that caps both the depth and breadth of analysis, and ultimately, the alpha institutional investors can generate.
Rethinking Alpha: Speed, Depth, and Scale
What if you could ask questions like:
- “Which industries with EV/EBIT under 10x are growing revenue north of 15%?”
- “What are the common margin trends across mid-cap retailers with China exposure?”
- “Which management teams are consistently optimistic on pricing power but missing guidance?”
And get high-quality answers in minutes, not weeks?
Today’s tools fall short:
- Alt-data platforms often surface raw signals (“search spikes,” “foot traffic surges”) without answering: So what?
- Generic LLMs, even GPT-4 Turbo with long context, stumble on domain-specific reasoning — failing 81% of benchmark finance questions.
- Even best-in-class open-source models still miss the nuance of sector-specific metrics, KPIs, and investor workflows.
What’s needed is not another dashboard or another chat interface.
It’s a domain-native AI analyst — one that understands how investors think, what they read, and how they decide.
Matterfact: Your AI-Powered Research Team
Matterfact is building exactly that.
Founded by Ashutosh Agarwal (ex-Nomura, WorldQuant, Google) and Vishal Verma (ex-Apple Siri, Google Gemini), Daniel Chen (ex-Morgan Stanley, Juniper Square). Matterfact is reimagining the investment research stack from the ground up — blending deep AI expertise with on-the-ground understanding of analyst pain points.
At its core, Matterfact is a purpose-built financial intelligence engine that:
- Ingests SEC filings, earnings calls, press releases, and market data in real-time
- Parses, ranks, and tags key metrics, management commentary, and market shifts
- Surfaces insights via interfaces analysts actually use — Chat UI, Bulk Analyzer, and the upcoming Model Bench
What makes it stand out?
→ Accuracy at scale:
Matterfact achieves 98%+ accuracy on FinanceBench, making it one of the most performant financial LLMs publicly benchmarked.
→ Workflow-native design:
Early users describe Matterfact not as a chatbot, but as “a sharp junior analyst who mimic’s our way of doing things.” The workflow interface allows users to codify their firm’s investment research process into logical flowcharts. The Screener tool looks at sectors cross-sectionally in seconds.
→ Real-time intelligence:
Unlike static tools, Matterfact continuously ingests and updates data. They access earnings transcripts live so investors can update their models right away. A surprise regulatory filing? Matterfact parses it, contextualizes its impact across your portfolio, and flags the outliers worth watching.
→ Measurable impact:
Early users are seeing 50–70% reductions in time spent on sector overviews and deep dives, freeing bandwidth to think more strategically.
Timing: Two Exploding Tailwinds
Matterfact is riding two secular surges:
- AI in Asset Management – Forecasted to grow to $48B by 2031 at a 34% CAGR
- Alternative Data in Finance – Set to reach $137B by 2030, growing at 53% CAGR
Despite these tailwinds, incumbents remain mostly data-centric and siloed. Bloomberg, FactSet, and legacy tools still treat AI as an add-on, not the core. Matterfact flips this: AI-first, insight-led, workflow-native.
Beyond Speed: A Vision to Reshape Finance
Matterfact isn’t just building tools — they’re reshaping the infrastructure of investment intelligence.
Their long-term roadmap is audacious, but deeply logical:
- Act I: The AI Research Analyst: Replace the drudgery of manual research legwork with a seamless AI teammate. In this phase (already underway), any investment firm, large or small, can plug into Matterfact and instantly equip their team with a tireless junior analyst who has read everything and can answer questions on demand. This democratizes a capability that was once limited to firms with huge analyst pools or expensive expert networks.
- Act II: Becoming system of record for Research: As more and more of firm’s investment process lives on Matterfact, more and more of their research process gets embedded inside it. Matterfact moves from reducing overhead to becoming central to how the research team operates.
- Act III: Owning the Long Tail of Research: As the platform ingests more data and user feedback, Matterfact can become the go-to source for high-quality research on the thousands of under-covered companies out there. Think about all the mid-cap and small-cap firms that get little love from Wall Street – Matterfact can produce on-the-fly “sell-side quality” coverage for those, leveling the informational playing field. An investor in, say, a $2 billion market cap company could ask detailed questions and get insights as rich as if a top analyst had spent a month researching it. This has implications not only for investors but for the companies themselves, who struggle to get attention from big banks. Matterfact could help surface hidden gems systematically.
In summary, Matterfact is positioning itself to go from analyst augmentation → to owning market knowledge → to redefining how capital is allocated and trades are executed. This spans the full stack of an institutional investor’s workflow, not just the initial research piece. For us, that means the upside isn’t just in making research more efficient; it’s in potentially reshaping the infrastructure of investment intelligence in the coming decade.
At z21 Ventures, we back founders solving hairy, high-pain problems in large, shifting markets.
We believe that over the next decade, every investment team will have an AI teammate. Matterfact is building the one that thinks like a financial analyst, not a search engine.
We’re excited to be on this journey with them.