Filings and earnings eat the week
Analysts spend days hand-parsing 10-Ks, 10-Qs and call transcripts before a single number reaches a model. The thinking starts only after the grunt work ends.
Your autonomous research desk
AI analyst agents that turn a week of research into ten minutes.
Quanova deploys autonomous research agents that ingest filings, news, market and alt-data, run quantitative analysis, backtest hypotheses, and deliver cited, audit-ready memos — with every number traceable to its source.
Built for research teams, hedge funds, and analysts.
NDX · daily
+2.41%
30d90d1Y5Y
RESEARCH MEMO
AUDIT-READY
Margin durability screen — semis
- Gross margin expanded 310 bps YoY, led by datacenter mix shift.1
- Inventory days fell to 118 from 142 quarter-over-quarter.2
- Backtested margin-momentum factor: 5Y info ratio 0.71.3
1 10-Q, p.14 2 10-Q, p.22 3 Backtest #4471
FILINGS
EARNINGS CALLS
MARKET DATA
ALT-DATA
NEWS
FUNDAMENTALS
The problem
Research doesn’t scale with headcount.
A fundamental thesis is gated by how fast a person can read, model, and footnote. Quanova removes that ceiling without removing the rigor.
Backtests are slow and brittle
Standing up a clean dataset, encoding a signal, and validating it across regimes is a multi-day loop. Most ideas die in the queue, not on the merits.
No traceability when it counts
When compliance or an IC asks “where did this number come from?”, the answer lives in someone’s spreadsheet history — not an auditable trail.
Product
An autonomous research desk, end to end.
Eight capabilities that take a question from raw sources to a cited, defensible memo — with the math and the citations attached.
Autonomous research agents
Agents read earnings, filings, news and fundamentals, decompose a question into sub-tasks, and assemble findings without hand-holding.
Quant analysis & backtesting
Factor construction, regressions, and historical backtests with reproducible parameters your quant team can re-run line for line.
Alt-data ingestion
Connect proprietary, vendor and alt-data feeds. Agents normalize, join and reason over them alongside market and fundamental data.
A citation on every claim
Each statement carries a footnote that resolves to the exact document, page, table or data point. No claim without a source.
Compliance audit trail
Every query, source touched and output is logged immutably. Reconstruct exactly how a conclusion was reached, months later.
Screeners & signal alerts
Define a screen once and let agents monitor the universe. Get alerted when a name crosses your criteria or a signal flips.
Natural-language market queries
Ask in plain English. “Which mid-cap industrials beat on margins but missed on revenue last quarter?” — and get a cited answer.
Exportable memos & dashboards
Export to PDF, Word, or a live dashboard. Citations and the audit trail travel with the document, ready for the IC.
How it works
From a question to an audit-ready memo.
Four steps. You ask in plain language; the agents do the gathering, the math, and the footnoting.
-
1
Ask a research question
Pose it in natural language, scoped to a name, sector or universe. No query language, no rigid templates.
-
2
Agents gather & cite sources
Filings, transcripts, news and your connected datasets are pulled, read and attributed — every fact tied to its origin.
-
3
Run quant analysis & backtests
Agents build the factors, run the regressions and validate across regimes, returning reproducible methodology.
-
4
Deliver an audit-ready memo
A cited draft lands in minutes — with the data, code parameters and audit trail attached for review.
ANSWER Two of three names expanded datacenter gross margin YoY; one shows rising inventory days.
- AVGO margin +310 bps YoY; inventory days 118.1
- NVDA margin +180 bps YoY; inventory days 96.2
- MRVL margin −40 bps YoY; inventory days 161 ↑.3
1–3 Latest 10-Q filings · methodology & audit trail attached
Use cases
One desk, many mandates.
Hedge funds & fundamental research
Deep-dive a thesis, build the comp set, and pressure-test the bear case — cited and ready for the PM in an afternoon.
Quant signal research
Prototype factors and signals against history, validate across regimes, and hand the quant team reproducible backtests.
Equity research & IR
Produce coverage notes and earnings recaps with traceable footnotes that survive editorial and compliance review.
Wealth & advisory
Answer client questions on holdings and markets with cited, on-brand summaries — documented for suitability files.
0+
filings parsed / minute
0%
of claims cited to source
0 min
typical run, not days
0+
data sources connected
Figures shown are illustrative.
Trust & compliance
Defensible by design.
Research that has to hold up in front of an investment committee, an auditor, or a regulator — so traceability and control are first-class, not bolted on.
- SOC 2-minded controls and processes
- Encryption in transit and at rest
- Role-based access & SSO/SAML
- Configurable data residency
- Immutable, queryable audit trail
- Your data is never used to train shared models
Pricing
Plans that scale with the desk.
Transparent tiers for an individual analyst up to a regulated enterprise. Annual billing; contact us for usage details.
Analyst
$1,200/ seat / mo
For individual analysts running cited research.
- Autonomous research agents
- Natural-language market queries
- Cited memos & PDF export
- Standard data sources
- Single user
Desk
$4,800/ mo
For research teams that share screens and signals.
- Everything in Analyst, plus
- Quant analysis & backtesting
- Screeners & signal alerts
- Alt-data & custom connectors
- Shared workspaces (up to 10 seats)
- Audit trail & team controls
Enterprise
Custom
For regulated firms with compliance requirements.
- Everything in Desk, plus
- SSO/SAML & SCIM
- Configurable data residency
- Dedicated environment & SLA
- Compliance & audit support
- Solutions engineering
From the desk
What research teams tell us.
“A first-pass memo that used to take an analyst three days lands in minutes — and every figure is already footnoted to the filing. That changed how we staff coverage.”
“The reproducible backtests are the real unlock. Our quants re-run the methodology line for line instead of arguing about a black box.”
“When compliance asks where a number came from, the audit trail just answers it. That alone justified the rollout.”
Quotes are illustrative and composite.
Founder
Vid Homšak
Founder & CEO
Vid is a physicist (University of Oxford) who built gradient-boosted models for particle-decay classification and ran algorithmic crypto portfolios on live exchange APIs. He kept seeing the same gap: brilliant analysts lose their weeks to data-wrangling and grunt work. Quanova is built on a simple conviction — a junior quant’s week of research should take an agent ten minutes, with every number traceable to its source.
FAQ
Questions, answered.
Is Quanova investment advice?
No. Quanova is a research and analytics tool. It produces analysis, citations and backtests to support your own decision-making. It does not provide investment advice, recommendations, or solicitations to buy or sell any security.
Where do the agents get their data?
Agents ingest public filings (10-K, 10-Q, 8-K), earnings transcripts, news, market and reference data, and any alt-data or proprietary datasets you connect. Every claim is linked back to its underlying source.
How does Quanova cite its sources?
Each factual statement in a memo carries a footnote marker that resolves to the exact document, page, table or data point used. Citations are preserved in the audit trail and travel with exported memos.
Can I run my own backtests and quant models?
Yes. Agents can run factor analysis, signal construction and historical backtests on connected datasets, and return reproducible methodology, parameters and results that your quant team can re-run.
How does Quanova handle security and compliance?
Quanova is built SOC 2-minded with encryption in transit and at rest, role-based access, configurable data residency, and a complete audit trail of every query, source and output for compliance review.
How fast is a typical research run?
Most research questions resolve in minutes rather than days. A multi-company comparison or a filing deep-dive that would take a junior analyst a week typically returns a cited draft memo in under ten minutes.
Give your desk an autonomous research team.
Request access or book a walkthrough. We’ll show you a live, cited research run on a name you care about.
Contact usFor research teams, hedge funds and analysts. No spam.