# Databuddy

Aiolum story: https://aiolum.com/story/databuddy/ · added 2026-11-05 · updated 2026-11-05

**In short.** Databuddy (Product analytics with an AI analyst, USA): Solo founder in YC Fall 2026 with paid plans live, November 2026. Evidence grade E: no revenue figure was checked. Team: 1 founder; Y Combinator lists a team size of 1 in November 2026. Launched 2025; still running.

Product analytics with an AI analyst · USA · AI-native · solo founder · still running · group: Documents, data and slides

- Outcome: Solo founder in YC Fall 2026 with paid plans live, November 2026 (evidence grade of the headline number: E; no revenue figure)
- Launch: 2025; months from launch to the outcome: 16
- Team: 1 founder; Y Combinator lists a team size of 1 in November 2026
- Outside funding: Y Combinator, Fall 2026 batch (amount not stated in the sources)
- Main factor: An AI analyst that shows its query
- Meters, 1 to 4: evidence strength 1, copyable pattern 2, capital needed 2
- Tags: AI-native, Solo founder, Funded, Running, B2B, Analytics

Links:

- Visit databuddy.cc: https://www.databuddy.cc/

**The product.** Website and product analytics from one cookieless script: visitors, sessions, custom events, funnels, goals, revenue from a connected Stripe account, error tracking, Web Vitals, feature flags and uptime checks. A built-in AI analyst called Databunny answers questions in plain words and shows the database query behind each answer. On the two top plans it also checks the metrics every day or week and posts real changes to Slack.

**Why it worked.** Its founder argues that collecting data is easy and understanding it is not: early startups use one tool for traffic, another for events, another for errors and another for revenue, and nobody on the team writes SQL. Databuddy puts the data in one place and lets an AI write the queries, then shows each query so the user can check it.

Evidence grades: A = disclosed by a public company, a filing or a payment processor, or verified by one; B = two or more independent reports agree, or a journalist confirmed it; C = a founder claim that no one else verified; D = reported, not verified: one founder claim or a third-party estimate with no second source; E = no revenue figure found, so the headline is a non-money fact and the other facts carry their own grades; calc = computed on this page from graded facts.

## Summary

Databuddy is product analytics, software that records what visitors and users do on a website or app (page views, sign-ups, clicks, purchases) and turns it into charts. Its tracker is cookieless: it does not store a cookie, a small file that identifies the browser, and the company says it avoids fingerprinting by default. The AI analyst, Databunny, turns a plain question into SQL, the language used to query a database, runs it, and shows the query next to the answer. Developers can also reach the data through an MCP server, a standard interface that lets AI coding agents such as Claude Code or Cursor call a service's tools.

Issa Nassar founded it alone. In the Y Combinator launch post of 5 November 2026 he writes that he started building it in Palestine and now builds it full time in the Fall 2026 batch. Y Combinator lists the company with a founding year of 2025, a team size of 1, San Francisco as its location, and the one-line description "AI Data Analyst for fast-growing teams".

No revenue figure was found. The pricing page lists a free plan up to 10,000 events a month and paid plans from US$9.99 to US$799 a month; the founder's traction claims are a Product Hunt ranking and views of an X post, which this page could not check.

## Pricing

Databuddy prices on 5 November 2026, from its pricing page and its public pricing endpoint (grade C)

| Plan | Per month, billed yearly (US$) | Per month, billed monthly (US$) | Limit |
|---|---|---|---|
| Free |  | 0 | 10,000 events a month, 10 AI credits |
| Hobby |  | 9.99 | 30,000 events, 20 AI credits plus 1 a day |
| Pro |  | 49.99 | 1 million events, 350 AI credits plus 5 a day |
| Business |  | 299 | 2 million events, 100 investigations, 1,500 AI credits |
| Scale |  | 799 | 6 million events, 250 investigations, 3,000 AI credits, SSO |
| Enterprise |  | custom | custom volume, security review, SLA |

All plans are billed monthly; the pricing FAQ says there is no annual billing. Events above the allowance cost US$0.035 per 1,000 on Hobby and Pro, with lower rates at higher volumes. Business and Scale are called early-user prices, locked while the customer keeps the plan, and extra investigations cost US$1 each. Free stops collecting events at 10,000 a month.

## How much revenue does Databuddy make? The numbers, graded

A one-person company in Y Combinator's Fall 2026 batch with five public paid plans and no published revenue figure

| Fact | Value | Grade | Note | Sources |
|---|---|---|---|---|
| Revenue | not published | E | No payment-processor profile, founder figure or filing was found. The launch post gives no revenue. | [2] |
| Founder and team | Issa Nassar, solo founder; team size 1 | C | Launch post: "I'm a solo founder." Y Combinator's company page lists team_size 1 and one founder. | [1, 2] |
| Founded | 2025 | C | Y Combinator company page, year_founded 2025. The month is not stated in any source read. | [1] |
| Accelerator | Y Combinator, Fall 2026 | C | Y Combinator company page and launch post, 5 November 2026. The investment amount is not stated in the sources read. | [1, 2] |
| Location | San Francisco; building started in Palestine | C | Y Combinator lists San Francisco, US. Launch post: "I started building Databuddy in Palestine". | [1, 2] |
| How Y Combinator describes it | "AI Data Analyst for fast-growing teams" | C | The one-line description on the Y Combinator company page. The pricing page title is "Pricing: an AI analyst for your product". | [1, 4] |
| Cheapest paid plan | US$9.99 a month | C | Pricing page and the public pricing endpoint, read 5 November 2026. Free up to 10,000 events a month. | [4] |
| Plans with AI investigations | US$299 and US$799 a month | C | Business and Scale, with 100 and 250 scheduled AI investigations a month. The pricing page compares them with a data analyst at "around $8,000 a month", its own estimate. | [4] |
| Traction claimed in the launch post | #5 Product of the Day on Product Hunt; Y Combinator announcement on X "passed 800K views" | C | Founder's claims, 5 November 2026. The Product Hunt product page answered 404 and X needs a login, so neither was checked. | [2] |
| AI model named by the company | Claude Sonnet, for AI insights | C | Changelog entry of May 2026: a length limit was widened to match "what Claude-Sonnet routinely writes". The data policy says AI requests go through Vercel's AI Gateway and "The provider depends on the configured model." | [7, 8] |
| Privacy design | no cookies, no fingerprinting by default | C | About page: "Databuddy avoids cookies and fingerprinting by default". The FAQ adds that the tracker uses browser storage and that "Cookieless does not automatically mean consent-free." | [3, 6] |
| Self-hosting | not offered | C | Home page FAQ: "Can I self-host Databuddy? No." The GitHub organization lists side projects; no main Databuddy repository was public on 5 November 2026 (the path answered 404). | [3, 9] |
| Time from founding year to the launch post | about 16 months | calc | Calculated: from mid-2025, the middle of the founding year, to November 2026. The true value lies between 11 and 22 months. | [1, 2] |

How to read this table: a grade C value is a statement by the company or its founder, on its own pages or its Y Combinator pages. A "calc" value was computed on this page. No value is grade A or B, and no revenue figure exists.

## When did Databuddy launch, and how did it grow? Timeline

From a solo project started in Palestine to a Y Combinator launch, as the company's own pages date it. Each entry carries its grade.

- **2025** (grade C [1, 2]): Issa Nassar starts building Databuddy alone in Palestine. Y Combinator lists 2025 as the founding year.
- **April 2026** (grade C [7]): The changelog records a bot-signature database of more than 970 crawlers and cost controls for the AI agent.
- **May 2026** (grade C [7]): A changelog entry on AI insights names Claude Sonnet as the model whose summaries the limits were sized for.
- **October 2026** (grade C [7]): Weekly changelog entries continue, with most entries signed by the founder's handle.
- **5 November 2026** (grade C [2]): Launch post on Y Combinator for the Fall 2026 batch: free up to 10,000 events a month, then from US$9.99, as the post states.

## Why did Databuddy work?

### An AI analyst that shows its query

Databunny answers a plain question, then shows the SQL it ran, so a founder can check the number instead of trusting it. The FAQ calls this the answer to "Can I trust the AI answers?"

### One script for several tools

Traffic, events, funnels, errors, Web Vitals, feature flags and revenue come from one install. The launch post lists Google Analytics, PostHog or Mixpanel, Sentry and Stripe as the separate tools it replaces.

### Priced on the AI work, not only on events

The low plans sell analytics with a few AI credits. The pricing page sells the two top plans as scheduled AI investigations that post findings to Slack, and compares them with the cost of a hired analyst.

### Built for coding agents

An MCP server, an llms.txt file and an OpenAPI description let agents such as Claude Code or Cursor read the analytics and create funnels or goals from the editor.

## How it was built

These details come from the company's changelog and data policy. They are grade C.

- Analytics events and revenue are stored in ClickHouse, a column database built for fast counts over many rows; webhook receipts are kept in PostgreSQL.
- Hosting is split across Hetzner, Unkey, Vercel, Bunny.net for the tracker files, Cloudflare at the network edge and PlanetScale for account data.
- AI requests go through Vercel's AI Gateway to the configured model provider; the changelog names Claude Sonnet for AI insights.
- Stripe takes payments and Autumn handles subscriptions, entitlements and metered usage; Resend sends email.
- The changelog is published weekly, and most entries name the founder's handle as author.

Sources: [7, 8]

## Who tried this before Databuddy, and why did they fail?

Three companies in our database of failed startups promised analytics insights to teams without a data team, or predictions from product data. The values below are from that database and are not independently verified.

The failed startups come from our database of dead startups, retold here in our own words; Its entries are AI-assisted summaries of public sources, so they carry no evidence grade.

### SumAll (2011 to 2019; funding US$15M; cause: Competition)

- Promise: Marketing and sales data from many platforms in one dashboard for small businesses without a large team.
- Why it died: It could not stand apart from Google Analytics, Tableau and free tools, and its price did not match the value small businesses saw.
- Record: 

### Preact (2012 to 2018; funding US$4.7M; cause: Competition)

- Promise: Predict which SaaS customers would churn from their product usage.
- Why it died: Larger customer-success platforms offered similar predictions inside a fuller product.
- Record: 

### Toplyne (2021 to 2024; funding US$15M; cause: No Market Need)

- Promise: Use product data and AI to find which free users would pay.
- Why it died: Budgets were cut after 2022, and the gain from better lead scoring was hard to prove.
- Record: 

## What changed

SumAll and Preact sold insight from data to teams that had no analyst, and both lost to larger tools that offered the same charts or predictions. Toplyne added AI predictions on top of existing analytics and could not prove the return. Databuddy is the same promise: answers without a data team.

Two things differ. A language model can now write the query for any question, not only for the reports a vendor designed, and Databuddy shows that query so the answer can be checked. And one founder with managed services (ClickHouse, Vercel, Stripe, Autumn) can run the whole stack, so the company does not need the funding the three predecessors raised before it finds out whether people pay.

## Can one person copy Databuddy now?

**Verdict: The product, hard. The idea, yes.**

Analytics is a crowded field with large, funded players, and Databuddy combines many features that took one founder over a year to build. The narrower idea can be copied: put an AI analyst that shows its work on top of data a customer already has.

1. Pick one data source that small teams already collect but do not read (payments, support tickets, ad accounts).
2. Let a model write the query, and always show the query next to the answer.
3. Sell scheduled checks that post only real changes to the team's chat, priced against the cost of a hired analyst.
4. Offer an MCP server so coding agents can read and change the setup.

**What is still open.** AI analysts for one data source each: Stripe revenue, app-store reviews, ad spend, support queues, or open-source package downloads.

## What could go wrong for Databuddy?

- **What is not verified.** No revenue figure was found. This page looked for a TrustMRR or other payment-processor profile, a founder interview with figures and a filing, and found none. The traction claims (#5 Product of the Day on Product Hunt, more than 800,000 views of an X post) are the founder's own, from the launch post of 5 November 2026; the Product Hunt product page answered 404 and X needs a login. The home page shows "Used by developers at" OpenAI, Vercel, Supabase, Upstash and CodeRabbit with no source for the claim. One quoted customer is described as "Context.dev" in the launch post and as "Founder, Rentmyheader" on the home page, and context.dev also appears in the data policy as a supplier for AI investigations. A reader could look for a revenue figure in a later Y Combinator Demo Day post or a payment-processor profile.
- **Competition from large analytics tools.** Databuddy names its competitors itself: its home page has comparison pages for Google Analytics, PostHog, Plausible, Mixpanel and Umami, and the launch post invites users who are "Switching from Google Analytics, PostHog or Plausible". Those products are older and larger, and the larger ones can add an AI analyst of their own. This page did not read their pricing.
- **Dependence on model providers.** The AI analyst runs on outside models. The data policy says AI requests go through Vercel's AI Gateway and that "The provider depends on the configured model", and the changelog names Claude Sonnet. Model prices or limits set by the provider change the cost of every AI credit and investigation the plans include.
- **Privacy claims and their limits.** The company says it avoids cookies and fingerprinting by default, and its own FAQ adds that the tracker uses browser storage and that cookieless does not mean consent-free. Whether a site needs a consent banner depends on its own setup and local law.

## Sources

1. Y Combinator company page: Databuddy — https://www.ycombinator.com/companies/databuddy (grade C). Read in full, 5 November 2026 (the page's embedded data). Source for the one-line description, founding year, team size, location, batch and founder.
2. Y Combinator launch post, 5 November 2026: Databuddy: Product analytics for founders without a data team — https://www.ycombinator.com/launches/WsC-databuddy-product-analytics-for-founders-without-a-data-team (grade C). Read in full. Founder's post. Source for the problem statement, the solo-founder statement, the start in Palestine, the traction claims and the price summary.
3. Databuddy home page — https://www.databuddy.cc/ (grade C). Read in full, 5 November 2026, including the FAQ answers in the page data. Source for the features, the self-hosting answer, the cookie answer and the customer quotes.
4. Databuddy pricing page and public pricing endpoint — https://www.databuddy.cc/pricing (grade C). Read in full, 5 November 2026, with https://www.databuddy.cc/api/pricing (allowed by robots.txt) for the exact allowances and overage rates.
5. Databuddy: Databunny, the AI analyst — https://www.databuddy.cc/databunny (grade C). Read in full. Source for how chat and scheduled investigations work.
6. Databuddy: About — https://www.databuddy.cc/about (grade C). Read in full. Source for the cookie and fingerprinting statement and the agent interfaces.
7. Databuddy changelog — https://www.databuddy.cc/changelog (grade C). Read the entries from March to October 2026 shown on the page. Source for the model name, the bot database and the authorship of entries.
8. Databuddy data policy — https://www.databuddy.cc/data-policy (grade C). Read in full. Source for the storage, hosting, billing and AI providers.
9. GitHub organization: databuddy-analytics — https://github.com/databuddy-analytics (grade C). Read the repository list, 5 November 2026. The path /databuddy-analytics/Databuddy answered 404.
10. Database of failed startups, local export — /method/#dead (grade C). Local export. AI-written summaries, not independently verified. Source for the failed predecessors.

Contact: aiolum@livyden.com. Run Databuddy? Send corrections or a verified figure to aiolum@livyden.com (subject: Correction: Databuddy).