Skip to content
MakerGain
← Back to project

He gave AI conversations a tree for the moments they need a detour

Edited by MakerGain

Slashspace handles the urge to pursue a side question while keeping the main thread intact when reading a long document: any answer can branch into its own node while the main conversation stays in place. Founder Praneeth Pike moved the product from a web subscription to a one-time desktop purchase, using BYOK to put model-call costs on users’ own keys and server costs on their computers.

Suppose you are reading a thirty-page PDF with AI. On page three, a term makes you want to ask what it means in this document. Follow that question and the main thread disappears up the scroll; finding the original sentence takes time. Open another window and you have to explain the context again.

Slashspace handles precisely that move. Branch from any answer on the canvas: the side conversation grows into its own node while the main thread stays in place. Its website puts it directly: branch from any response and stop scrolling back. [1]

First, what is it? A desktop app that brings PDFs, YouTube videos, webpages, and notes onto one canvas, with typed or spoken questions. The website says chats read complete sources rather than a search index. Users can choose ChatGPT, Claude, Gemini, or local models, ask side by side on the same canvas, and switch with one click.

Image: from the product website[1], obtained on 2026-11-18. Original official illustration, not an independently tested result.

On May 19, 2026, RabbitHoles AI announced its rename to Slashspace AI. The announcement is candid about why: models had become stronger and users more experienced, so branching conversations alone no longer covered the opportunity. More agentic features would follow, with a stronger canvas-first approach; canvases were also renamed spaces. Rename announcement[2] The website's current Orchestrator mode follows that direction: plan work, create and connect nodes, delegate tasks to subagents, and bring back results. It lists Claude Code, OpenAI, OpenRouter, Grok, and Gemini as entry points, and can connect tools such as Notion, GitHub, and Gmail through Composio/MCP.

Changing three things at once

Founder Praneeth Pike explained the desktop app and one-time purchase in a Starter Story interview. The route was not especially pretty. He began with the standard setup: subscriptions, a free trial, and a web app, without meaningful traction. His judgment was that the target audience consisted of power users already paying for several AI subscriptions, while he had no substantial personal brand to draw on. [3]

He changed three things: move the web app to the desktop, shifting server costs to users' computers; replace subscriptions with a one-time purchase; and use BYOK to put AI-call costs onto users' own keys. He was direct in the interview: a one-time purchase meant he could not absorb AI usage or server costs himself, and “the simplest way is to use the user's computer.”

BYOK is only one route, however. The official billing policy describes another: both the $89 annual plan and the $249 lifetime version include a one-time allocation of 500 credits. Subscription renewal neither resets nor grants them again. 1 credit equals 1 cent of usage on hosted models and tools; additional usage is sold at US$11 for US$10 of usage. Bringing an API key, running a coding-agent app through one's own Claude/ChatGPT subscription, or using a local model consumes no credits. The tradeoff includes failures: if a personal key expires or is rate-limited, the request fails. It does not automatically fall back to hosted credits unless the user switches to a hosted model. [4]

What the money looks like

The interview gives roughly $80K gross revenue and $75K profit in the six months after switching to one-time purchases. He said refunds and other items had been deducted. There were about 200 paying users, including 60 lifetime buyers. The initial price was $29 with fewer than 100 users, rising to about $49 after passing 100. He also said he planned to raise the lifetime price further and move toward subscriptions; there were about 1,200 users at the interview. He described day-to-day running costs as almost zero: about $50 a month for landing-page hosting and $200 a month for Cursor, used only to develop new features rather than maintain the existing product. Profit is his account, without a complete cost statement or audit.

His acquisition story is similarly plain. First redo the landing page, aiming the H1 and copy throughout the page at power AI users. Then join Twitter discussions about AI interaction, without pitching directly, only hinting at a relevant product. At the interview, the website had accumulated about 38,000 unique visitors. Channel-level conversion data is undisclosed. That figure establishes visits.

Public data represents a different point in time. A TrustMRR snapshot synchronized with Stripe at 2026-11-17T14:22:16.261Z shows MRR of $182, 20 active subscriptions, $2,024 revenue over the previous 30 days, and $132,327 cumulative revenue. It also records +77.1% MRR growth in the preceding 30 days. The public page does not break revenue down among subscriptions, lifetime purchases, and credits. [5]

The distinctions matter: MRR and actual receipts are different metrics; 20 active subscriptions do not equal all customers; and this is a single snapshot, not a trend line. The interview was in October 2025, the snapshot in November 2026. The two points cannot simply be subtracted, nor can 200 paying users be compared directly with 20 subscriptions. The snapshot's monthly revenue table runs from $5,103 in 2025-12 to $1,167 in 2026-11. November 2026 only runs through the November 17 sync, not a full month. These months show continuing receipts after the interview. Their mix of subscriptions, lifetime purchases, and credits is not visible.

The current official policy lists $89 a year or $249 for a one-time lifetime purchase, charged at payment, without a free trial. Either plan allows refund requests within 7 days of payment, deducting transaction fees, ending access, and removing included credits; canceling a subscription does not automatically refund it. The homepage's Yearly section has displayed both $89 and $43.61. It labels $43.61 as the annual price after code BF2026, ending December 1, while the official policy gives the standard $89 annual price. AppSumo has different terms: lifetime access at $59, $139, $189, or $289. The $59 tier lists 1 device, 100 local canvases, and 0 synced canvases. Synced canvases and cloud storage appear from Tier2 upward, with a 60-day refund period on the page. AppSumo page[6] The channels' refund periods must be kept separate.

Synced canvases reveal another unresolved point. The website lists Cloud sync as Coming soon and describes annual-plan benefits as “canvases saved on your own computer.” AppSumo already lists synced canvases as a selling point for Tier2 and above. Public pages do not reveal what purchasers have actually received; they only show the discrepancy. The same Coming soon list includes video generation, team sharing, and collaborative canvases.

Others are working here too

Google's Gemini Notebook also lets users upload multiple sources and ask questions. Answers cite direct quotations, text, and images from the sources; clicking a citation returns to its original context. Chat uses only source data, and creative tasks may be refused. Gemini Notebook help[7] The help page describes cited source-based Q&A without mentioning canvases or branches. It handles the part about controlling sources and making answers traceable.

A closer reminder is Kosmik. Its official notice announced a shutdown on 2026-05-31 and emails to users and subscribers starting on 2026-04-24 about data export and subscription cancellation/refunds. Kosmik notice[8] It does not explain the shutdown's cause. This predicts nothing about Slashspace; it is simply an exit in a neighboring direction.

What did he actually sell?

That is the question worth considering. The 500 credits have a clear usage rate of 1 cent per 1 credit, with more to buy when they run out. The remainder is the shape of a conversation: a main thread, branches that can grow at any point, and a map of one's own thinking. Shifting variable costs to users through BYOK and a desktop app lets an independent developer sell software outright. That is the smartest part of this business.

It is also the tightest constraint. Cloud sync and team collaboration on the Coming soon list mean working across devices and sharing among people, neither of which is simply a single-computer task. Public pages do not say whether video generation must use hosted models. He originally shifted server costs to users' computers. If cloud sync and team collaboration arrive, the public pages do not answer who will bear that delivery and its costs.

Continue reading

AI productivity
Klipy, the sales AI that waits for your approval

Klipy takes over follow-up after sales meetings. It captures customer conversations from email, meetings, and instant messaging, drafts follow-ups and CRM updates, and prepares for the next meeting. Follow-up messages go out only after the salesperson approves them. Founder Jung Hong Kim is a Korean serial entrepreneur who took on debt when COVID sank his retail venture. He met his cofounder while consulting and launched Klipy.ai in three months. It went live in November 2024 as an AI CRM that logged communications automatically. In June 2025, it shifted toward carrying out next steps for sales teams and tied billing to specific actions users approve.

AI productivity
The founder reported \$8,000 in 2024 revenue, 85% collected only once: NodeLand turns long reports into maps

Rio de Janeiro-based Gabriel Xará runs NodeLand as a side project. Canvas extracts material from long reports, video transcripts, or meeting notes and generates structured maps; people still decide which nodes to keep, whether relationships are correct, and who the map is for. Built in 2023 and launched in 2024, it acquired early users through platforms such as AppSumo. A January 2025 interview disclosed \$8,000 in 2024 revenue, 85% from lifetime licenses, with costs and profitability undisclosed. He had focused almost entirely on the product that year, with limited marketing.

AI productivity
After closing a product nobody wanted, they broke entrepreneurship into “the next step”

David Heikka and his brother Felix launched their first AI form product and found zero demand. They tried ads, landing-page changes, renaming, and added features, then closed it after six months. They went on to build Buildpad, breaking business creation into stages that indicate what to validate next. Its early and current billing mechanisms differ, and the product was renamed aicofounder before changing back to Buildpad.

AI productivity
The client owns the code and pays the AI bill: SpaceAI360 builds automation you can take with you

Founded by Shahdin Salman alone, SpaceAI360 builds portable AI workflows inside customers' own accounts. It handles parsing, OCR, field extraction, database checks, and ERP/CRM writes after a form, email, or file arrives, routing uncertain items to human review. Rather than generic AI consulting, he starts with the highest-friction bottlenecks, assigning deterministic steps to n8n and judgment steps to Claude.