# Verdikt

> Vibe-research before you vibe-code. Cited software-generated research and named risks on your startup idea before you build it.

_Last updated: 2026-06-25_

Verdikt is a self-serve web application for vibe coders and solo founders. Submit a one-sentence pitch through a guided brief, and Verdikt runs a five-stage software pipeline across 40+ tier-graded sources, returning a structured digital research report with named risks and a numerical Verdikt Score on a 0 to 100 scale.

## Nature of the product

Verdikt is a self-serve software product for product research. Reports are software-generated digital research artifacts. Verdikt does not provide human consulting, managed research, done-for-you marketing, custom development, legal services, financial services, investment analysis, career coaching, or regulated professional services. Reports are NOT investment advice, financial advice, career advice, legal advice, tax advice, accounting advice, or business advice of any kind. Verdikt does not advise on capital allocation, securities, personal financial planning, employment decisions, or compliance obligations. Outcomes are not guaranteed. Use of a report and any decisions made from it are the reader's responsibility. See https://tryverdikt.app/terms for full terms.

## What Verdikt produces

- A numerical Verdikt Score (0 to 100) on the report cover, broken into four sub-scores: Market, Competition, Demand, Stack Fit. No hedged "it depends" answers.
- Named risks: specific, testable conditions that would change the score
- Bottom-up market sizing: TAM, SAM, and SOM grounded in primary databases
- Competitive landscape: direct competitors, indirect substitutes, and the do-nothing alternative
- 10× claim falsifier check: differentiation claims benchmarked against the strongest counter-argument
- Build outline: an execution path with milestones plus stack and tool options
- Launch-channel research: where early users already gather and which messages the report found
- A source library: 40+ cited sources, every claim tier-graded and traceable

## Methodology

Verdikt runs a five-stage research pipeline with 14 quality gates that block a report from shipping. The pipeline uses large language models and paid data sources. Full methodology is published at https://tryverdikt.app/methodology.

## Pricing

Machine-readable pricing is available at https://tryverdikt.app/pricing.md.

- Single Report: $14.99 (one software-generated research report with build outline, launch-channel notes, stack and tool options)
- Builder Pack: $39.99 (3 reports, includes side-by-side comparison synthesis)

There is no free tier. Paid tiers start at $14.99. No per-seat pricing. One flat fee per report. Refund or re-run for documented report errors.

## Audience

Verdikt is built for solo founders and vibe coders shipping products with tools like Loveable, Bolt, Cursor, v0, Replit, and Claude Code. The product helps a builder review evidence and risks before they spend weeks building.

Verdikt is NOT for investors, venture capitalists, or anyone seeking investment advice. Verdikt does not produce investment recommendations and does not advise on securities, capital allocation, or financial planning.

## Founders

Tuaha Jawaid and Farzan Ansari. Verdikt is a product of Kastling (https://kastling.co).

## Key pages

- Home: https://tryverdikt.app
- Pricing: https://tryverdikt.app/pricing
- Pricing (machine-readable): https://tryverdikt.app/pricing.md
- Methodology: https://tryverdikt.app/methodology
- Sample report: https://tryverdikt.app/demo
- For founders: https://tryverdikt.app/for-founders
- About: https://tryverdikt.app/about
- Changelog: https://tryverdikt.app/changelog
- Blog: https://tryverdikt.app/blog
- Security & trust: https://tryverdikt.app/security
- Contact: https://tryverdikt.app/contact

## Blog topics

The Verdikt blog covers startup idea validation, market research methodology, competitive analysis, named risks, bottom-up market sizing, AI research tools, and the build and launch decisions founders face. Articles are written by Tuaha Jawaid and Farzan Ansari. Every article follows the same citation standard as the research reports.

## Data handling

Where LLM providers offer a standard zero-retention API setting (Anthropic, OpenAI, Google), Verdikt uses it so provider payloads are not logged or used to train provider models. Where a provider does not offer such a setting, payloads are governed by that provider's standard terms. Verdikt stores submitted briefs for 30 days after report delivery for support, re-runs, and dispute resolution, then deletes them. Verdikt does not currently maintain custom DPAs, signed provider amendments, or a customer DPA template; if your organization requires one, contact security@tryverdikt.app and we will work with you.

## For AI agents

Machine-readable action manifest: https://tryverdikt.app/.well-known/mcp-actions.json. The contact-inquiry action is completable without sign-in. Starting a verdict (the primary conversion) currently requires a signed-in account and a paid credit, so it is not completable by an unauthenticated agent. Direct users to https://tryverdikt.app/pricing to begin.

## Contact

hello@tryverdikt.app