Branching research workspace with cited reports

For

Researchers, analysts and research teams exploring a topic through branching AI conversations

Solves

Branching AI conversations scatter findings across threads, so sources, context and reusable notes are lost between sessions.

Delivers

Cited, versioned research report with a linked evidence graph

Built in

about 5 weeks of creation time, MVP in 6 days

Investment

$14,500 for the MVP, $49,500 for the full product

Run it

Inside your business, or as part of your offer to clients

Keep branching exploration and its evidence in one owned workspace.

1. Start branching conversations from one topic.
2. Map ideas and branches as a visual graph.
3. Arrange chats, notes and projects on an infinite canvas.
4. Create child pages that inherit parent context and can be re-grounded with new sources.
5. Attach citations and timestamps to sources.
6. Link notes to the research they came from.
7. Continue research as linked chat nodes without restarting.
8. Switch between permitted AI models for tailored responses.
9. Save progress locally for continuity.
10. Store and link research and references in a local-first knowledge graph.
11. Publish, fork and import shared references.
12. Capture external research through web clipping.
13. Support real-time co-editing of documents and decks.
14. Retain shared memory across threads and projects.
15. Generate reports, visuals and presentations from one query.
16. Build structured learning decks with non-linear paths.
17. Delegate multi-step work to an agent with a large contextual scope.
18. Export to Markdown, PDF, DOCX or ZIP.
19. Compare the reviewed result with the recorded baseline and value assumptions.
20. Capture corrections and named-owner approval before consequential use.
21. Export a versioned cited, versioned research report with a linked evidence graph with source references and unresolved questions.

Everything these tools do, in one app

• Branching conversations Lets you follow multiple conversation paths from one discussion instead of restarting a single thread.Found in Innogath, Rabbitholes, DiveDeck.AI
• Visual idea graph Shows ideas and branches as a visual map so you can see how threads relate.Found in Innogath, Rabbitholes, Flowith Canvas
• Infinite canvas workspace Provides a large spatial workspace for arranging chats, ideas, and projects.Found in Rabbitholes, Flowith Canvas
• Structured report output Turns research into a readable, book-style report.Found in Innogath
• Source citations and timestamps Attaches citations and timestamps to sources so you can judge how current a claim is.Found in Innogath
• Branching pages with context Creates child pages that inherit parent context and can be re-grounded with new sources.Found in Innogath
• Linked notes Connects notes to research so follow-up work stays in context.Found in Innogath
• Node-based chat Continues research as linked chat nodes without starting over from a blank thread.Found in Innogath
• Export options Exports work in formats such as Markdown, PDF, DOCX, or ZIP for handoff and archiving.Found in Innogath
• Multi-model support Lets you switch between different AI models for tailored responses.Found in Rabbitholes
• Local data saving Saves progress locally for continuity and easy reference.Found in Rabbitholes, Subgrapher
• Local-first knowledge graph Stores and links research and references locally under the user's control.Found in Subgrapher
• Shareable references Lets users publish, fork, and import shared references.Found in Subgrapher
• Integrated organizer and mail Combines a mail-style client and personal organizer for time and event management.Found in Subgrapher
• Decentralized messaging and voting Supports messaging and community voting to help with discovery and curation.Found in Subgrapher
• Local model interaction Lets you interact with local models through a remote interface to assist reasoning.Found in Subgrapher
• Real-time co-editing Lets multiple people edit documents and slides together in real time.Found in Agnes AI
• Shared memory Retains context across threads and projects for long-term collaboration.Found in Agnes AI, Flowith Canvas
• Multi-agent content generation Generates reports, visuals, and presentations from a single query.Found in Agnes AI
• Cross-device collaboration Works across devices for live or asynchronous teamwork.Found in Agnes AI
• AI-generated content decks Builds structured learning decks from topics, concepts, or questions.Found in DiveDeck.AI
• Non-linear learning paths Allows branching into related subjects and exploring at your own depth.Found in DiveDeck.AI
• Multi-step agent tasks Delegates multi-step work to an agent that maintains a large contextual scope.Found in Flowith Canvas
• Web clipping Captures external research into the workspace.Found in Flowith Canvas
• Real-time sharing and commenting Supports sharing and commenting for collaborative work.Found in Flowith Canvas

What goes in, what comes out

The workflow

1. In
   Start with

   Permitted sources, conversation branches and reviewer notes

2. 1

   Confirm the buyer's problem and scope

3. 2

   Collect permitted sources

4. 3

   Conversation branches and reviewer notes

5. 4

   Then follow this sequence: 1

6. Out
   Finish with

   Cited, versioned research report with a linked evidence graph

AI does the heavy lifting, people stay in charge

Use AI to interpret permitted inputs, suggest structured mappings and generate candidate outputs for the three stated task modules. Use deterministic code for arithmetic, schema validation, hard constraints and reproducible tests. Review source-linked explanations and uncertainty before accepting results. One fixed citation format and permitted model set; final source checks and claims remain editorial. A model suggestion is never a verified fact, professional decision or authorization to act.

What your team sees

Primary screens: Topic and source intake, Branching canvas, Evidence graph, Report workspace, Review and export. Use a project gallery, a large spatial canvas for branches and nodes, and a right-hand panel for sources, citations, notes and comments. Let users compare branches side by side and re-ground a branch with new sources. Display draft, changes requested and approved states. Provide a client preview link with comments anchored to the relevant node or claim. Make the task-specific outcome cited, versioned research report with a linked evidence graph visible beside its evidence, review state and value baseline.

Accounts and administration

Project ownership, source versions, client comments, approval states, usage allowances, revision limits, download history and a rights record for supplied material. Add organization access boundaries, named reviewers, usage caps, data retention controls, export logs and explicit approval for external actions.

Integrations and data access

Author-owned sources, permitted databases and authorized interviews. Cloud storage, document import/export and publishing destinations. Start with file exchange and validate destination specifications before promising direct publishing. Start with authorized file exchange. Validate current provider access, usage rights and schema behavior before promising a connector.

We build with our own AI software development factory, so most implementations take days to a few weeks of creation time, not months. You see working software at every step, and exact timing depends on availability.

1. 1

   Scoping call

   Day 1

   Thirty minutes on your process, your data and how you want to run it: for your own team, or for your clients. You get a fixed scope and price for the MVP.

2. 2

   MVP

   6 days

   One buyer segment, one recurring use case; first modules: start branching conversations from one topic; map ideas and branches as a visual graph. Manual review in the loop. Built by our AI software factory.

3. 3

   Paid pilot

   7 days

   Accounts, roles, review states, audit trail and the first integration, hardened for two to three paying pilot customers.

4. 4

   Full product

   3 weeks

   Self-serve onboarding, billing, monitoring and the wider integration set.

5. 5

   Run and improve

   Monthly

   We host, monitor and improve it for a fixed monthly fee, or hand it over to your team. How the retainer works.

Why we start with an MVP

An MVP, or minimum viable product, is the smallest version that your users can actually work with. It is not a cheap version of the full solution. It is a test, built to answer the questions that decide whether the rest is worth building.

1. Pick the riskiest assumption. Here: will researchers, analysts and research teams exploring a topic through branching AI conversations use it to solve "branching AI conversations scatter findings across threads, so sources, context and reusable notes are lost between sessions"?
2. Build only what tests it. One team, one use case, a few core modules. People do the rest by hand for now.
3. Run a paid pilot. Agree quality and outcome thresholds before the pilot using this measure: Accepted report sections per research hour and unsupported claims found in review.
4. Measure, then decide. Track accepted report sections per research hour and unsupported claims found in review; accepted-output rate; material error rate; reviewer correction time; actual repeat purchase. Then expand, change course or stop, with evidence instead of opinions.

MVP scope for this solution. Pilot scope: One fixed citation format and permitted model set; final source checks and claims remain editorial. Implement one approved input format, a bounded representative case set and the first two task modules: start branching conversations from one topic; map ideas and branches as a visual graph. Support the third module with operator review: attach citations and timestamps to sources. Include source references, corrections, basic organization access, approval states, export and value measurement. Use managed operator assistance for unresolved exceptions. The cost estimate covers this narrow prototype, not unrestricted multi-tenant scale, complex production integrations, specialist certification or physical operations.

After the MVP. Once paid pilots prove usefulness, automate repeatable reviewed steps and add one verified source integration. Expand supported inputs and case volume only after new evaluation cases pass. Build reusable customer configurations and recurring value reports around cited, versioned research report with a linked evidence graph. Retain the explicit scope boundary: One fixed citation format and permitted model set; final source checks and claims remain editorial.

What the build depends on. Source upload and preview, asynchronous generation jobs, editable version history, reviewer access and tested export formats. High-fidelity research requires specialist source QA. Obtain representative authorized cases, baseline measurements, qualified reviewers and a buyer-side decision owner. Specific limitation: One fixed citation format and permitted model set; final source checks and claims remain editorial.

A planning range to start the conversation, not a quote. You pay per phase, so you can stop after the MVP.

Running costs per month

A rough indication of monthly hosting and AI model costs once it is live, not tested. Real costs depend on usage, file sizes and the models chosen.

Your brand, or this one

Run it under your own brand, or start from this concept style.

Selling it to your own clients: the go-to-market playbook

Pricing to test

Test a USD 300-1,500 fixed pilot for one defined research package. Offer a monthly production allowance after repeat demand. Quote complex multi-team or specialist research separately. These are test prices, not market benchmarks. Package the initial sale as one bounded cited, versioned research report with a linked evidence graph. Recurring fees must specify volume, review depth and integration support. For exchanges, test a disclosed coordination or successful-service fee rather than holding customer funds. Reprice only after measuring real delivery labor; platform-build cost is separate from a commercial pilot fee.

Message to test

Keep branching exploration and its evidence in one owned workspace. Demonstrate a concrete cited, versioned research report with a linked evidence graph using the buyer's approved example and show the baseline, corrections and actual delivery effort.

Where to find buyers

Researchers, analysts and research teams exploring a topic through branching AI conversations professional communities; specialist consultants serving this buyer; permissioned partner introductions; practical demonstrations at relevant trade or practitioner events.

Lead magnet

A reviewed sample cited, versioned research report with a linked evidence graph from a small authorized input set, with a transparent calculation of accepted report sections per research hour and unsupported claims found in review and no promised savings.

The first 30 days

1. Week 1: interview five researchers, analysts and research teams exploring a topic through branching AI conversations and inspect a recent example of branching AI conversations scatter findings across threads, so sources, context and reusable notes are lost between sessions.
2. Week 2: prepare a consented or synthetic demonstration of the three task modules.
3. Week 3: seek one bounded paid pilot with agreed baseline and acceptance criteria.
4. Week 4: measure accepted report sections per research hour and unsupported claims found in review, reviewer effort and repeat-purchase interest. This is a demand-validation plan, not a thirty-day full-product delivery promise.

Paid pilot

Agree quality and outcome thresholds before the pilot using this measure: Accepted report sections per research hour and unsupported claims found in review. Continue only if the buyer accepts the actual output, the intended job outcome improves without unacceptable errors, and measured delivery cost fits willingness to pay. Revise or stop if access is unavailable, qualified review cannot be provided, or apparent savings disappear after corrections and support. Use held-out cases when comparing model quality; use a properly reviewed comparison design before making causal claims. Record missing cases and negative results alongside successful outputs.

Success metrics

Accepted report sections per research hour and unsupported claims found in review; accepted-output rate; material error rate; reviewer correction time; actual repeat purchase.

Retention and expansion

Repeat the workflow when the buyer again needs cited, versioned research report with a linked evidence graph. Retain permissioned settings and reviewed examples, report realized value honestly, and sell increased volume or adjacent approved workflows only after contribution margin and quality remain acceptable.

Why clients would pick it

A reusable library of approved citation formats, source constraints and review examples, together with reliable delivery for a narrow research niche. Build a permissioned library of representative task cases, reviewer corrections and verified operating constraints for researchers, analysts and research teams exploring a topic through branching AI conversations. Repeatable delivery and useful integrations matter more than access to a base model.

Alternatives and positioning

Innogath, Rabbitholes, Subgrapher, Agnes AI, DiveDeck.AI and Flowith Canvas, plus generic chat tools and note apps. Compare this product with the buyer's present method on accepted report sections per research hour and unsupported claims found in review. Offer a bounded paid workflow instead of claiming broad autonomous expertise. Market uniqueness and competitor coverage are not verified.

Main delivery costs

Model calls, storage, reviewer hours, client revision rounds and licensed source material. Additional initial validation requires representative authorized sample preparation, buyer interviews, buyer-side evaluation and bounded validation of cited, versioned research report with a linked evidence graph. Track cost per accepted output, including correction work, unsuccessful cases and support.

Preserve source attribution, quotation accuracy and usage permissions. Named reviewers approve substantive claims and publication scope. One fixed citation format and permitted model set; final source checks and claims remain editorial. Keep all consequential actions under authorized human control and do not fabricate missing inputs, permissions, professional judgments or market evidence.