Stocknet
Features

The full toolkit for AI-run research.

Every Stocknet feature shares one workspace and one memory, so your agents compound on each other’s work instead of starting from scratch. Here’s what’s inside.

See it work
01 AI Agents

A specialist agent for every job.

Spin up autonomous agents that read, reason, and produce cited work. Each one is built for a task and draws on the same shared workspace context.

Run in parallel

Kick off research, monitoring, and screening jobs at once and let them work in the background.

Show their work

Every agent exposes its steps and sources, so you can audit how it reached an answer.

Hand off between tasks

A screening hit becomes a research brief in one click, carrying its context along.

agents · workspace 3 running
Equity Researcherreading 10-Q…
Idea Scoutscreening 412 names…
Portfolio Monitordone · 4 flags
Filing Readeridle
Earnings Trackerscheduled · Thu
02 Unified Memory

One memory that learns how you invest.

Every agent reads from and writes to a single workspace memory. It remembers your holdings, your thesis on each name, and the way you make decisions, then applies it to new work.

Learns your style

Picks up what you weight and what you avoid, and filters suggestions to match.

Shared across the team

Research done once becomes context for every agent and teammate after.

Guides new ideas

New cases arrive pre-aligned to your mandate, with the reasoning shown.

workspace-memory · ridgeline-capital
14
Positions
6
Live theses
41
Cited memos
Learned preference
Pass on names with net leverage > 3x. Screens now filter accordingly.
style: quality compoundersavoid: high leverage
03 First-Source Citations

Every number traces back to its document.

Stocknet pulls from primary sources and attaches the citation inline. Hover any figure to see the exact filing, page, and quoted passage. If it can’t be cited, it isn’t stated.

Primary sources only

Filings, transcripts, and official releases, ingested at the source.

Open the receipt

Each figure links to the document, page, and exact passage it came from.

No silent guessing

When the source doesn't support a claim, the agent says so.

Agent answer · NVDA Q1 FY26

Data-center revenue reached $39.1B, up 73% YoY, now 88% of total revenue1 source10-QNVIDIA Form 10-QSEC EDGAR · filed 2025-05-28 · p.4“Data Center revenue was $39.1 billion, up 73% from a year ago… and represented approximately 88% of total revenue.”

04 Portfolio Monitor

Wake up to only what moved your thesis.

The monitor watches every name you hold and triages the noise, surfacing filings, guidance changes, and price action that actually matter, each linked to its source.

Your real holdings

Tracks the positions in your workspace, not a generic index.

Thesis-aware triage

Filters events down to the ones that touch your reasons for owning.

Sourced alerts

Every flag links to the filing or release behind it.

portfolio-monitor · 14 positions live
CRWD · 8-K filed: CFO transition announced
sec edgar · 6m ago
META · raised FY capex guide to $72B
earnings call transcript
ASML · bookings below consensus
q3 press release
COST · monthly sales +7.1% y/y
company release
05 Filings & Transcripts

Primary documents, read at the source.

Agents ingest 10-Ks, 10-Qs, 8-Ks, and earnings transcripts directly and pull the figures that matter, each tagged to the page it came from. No scraping, no summaries of summaries.

Direct ingestion

Documents come straight from EDGAR and official investor sources.

Figure extraction

Key numbers are pulled with their context and page reference.

Always current

New filings are picked up as they hit the wire.

filing-reader · NVDA 10-Q
Item 2 · MD&A p.4
Data Center revenue was $39.1 billion, up 73% from a year ago, driven by demand for accelerated computing platforms, and represented approximately 88% of total revenue for the quarter.
DC rev $39.1BYoY +73%mix 88%
06 Research Terminal

One workspace for the whole research loop.

The terminal ties it together: prompt an agent, watch it reason over sources, get a cited answer, and keep the thread. Everything you and your agents produce stays in one place.

Conversational

Ask follow-ups against the same sourced context.

Structured output

Tables, memos, and figures, all cited and exportable.

Everything in context

Holdings, theses, and past work are one prompt away.

terminal · TSM cited
Thesis · auto-generated

Structural leader in advanced nodes; 3nm ramp and pricing power offset cyclical demand.

Key figures
Gross margin
59.6%
+2.1 pts q/q
HPC revenue
52%
of total
Capex FY
$42B
guided high

Your research desk, fully staffed by AI.

Start free, point an agent at your first idea, and watch it work, with every claim cited back to the source.

Back to overview