Live demo · MCP for Cursor · No signup

Market sizing you can
defend in the room.

Chronicle runs your market through a multi-agent pipeline that searches the web, papers, and news, scores every source for credibility, flags contradictions, and hands you a cited TAM/SAM/SOM report — not a paragraph you have to fact-check.

Numbers with receipts

The difference between an estimate and a defensible number.

ChatGPT gives you a number. Chronicle gives you a number you can source, stress-test, and stand behind in a partner meeting.

Every source scored

Each source gets a credibility score before it reaches your number. You see the average confidence up front — not a black box.

Contradictions surfaced

When sources disagree on a number, Chronicle flags it instead of silently averaging. You walk into the room knowing what's disputed.

Every figure cited

TAM/SAM/SOM with a clickable source behind each claim. When a partner asks "where's that from?", you have the link.

Verified case study

How an analytics firm shipped research 10× faster.

Manual research was taking 10+ hours per report; analysts spent most of that time aggregating data instead of generating insights. Chronicle’s multi-agent pipeline replaced the aggregation step end-to-end.

10×
faster research output
80%
cost reduction per report
0
manual citations — fully automated
“Their multi-agent research system transformed how we operate. Research that took days now takes minutes — and the quality is consistently better than manual work.”— Priya S., Head of Research, Analytics Firm
Inside the pipeline

Five specialized agents. One auditable trail.

Not a single LLM call. Each agent is picked for the specific cognitive load of its job — and you watch every step run live.

01 · Retriever

Pulls from web, papers, and news

Tavily, Perplexity, and arXiv in parallel. Sources scored before they hit your report.

02 · Enricher

Adds metadata + sentiment

Dates, author signals, source-type classification, sentiment polarity. Every claim gets context.

03 · Analyzer

Surfaces contradictions

Claude 3.5 Sonnet flags conflicts between sources.

04 · Insight

Builds the reasoning chain

GPT-4o turns scattered claims into hypotheses + trends.

05 · Reporter

Compiles a cited markdown report

Claude 3.5 Haiku assembles the deck-ready output. Drop your TAM/SAM/SOM straight into a slide, Notion, or a YC application.

How it works

No black box. Every step the agents take is visible, recorded, and replayable.

  1. 01

    Name your market

    Type a market, segment, or category. No prompt engineering, no API keys, no signup.

  2. 02

    Watch five agents fan out

    Web, papers, and news in parallel. Every source scored, every contradiction flagged, the reasoning chain assembled live.

  3. 03

    Get a cited sizing report

    TAM/SAM/SOM with sources, surfaced contradictions, and a confidence score. Markdown export — paste straight into your deck.

Built for the questions founders actually ask.

  • Pre-fundraise — size your market with numbers you can defend in a partner meeting.
  • Pre-launch — map the competitive landscape, find the gaps, find the prior art.
  • Customer discovery — surface what real users are saying about today’s solutions, with sources.
  • YC application — back every claim with a citation a partner can click. Or run research from Cursor via MCP while you draft.

Stop guessing your TAM.

Run your market through Chronicle. Free, open-source, and every number comes with a source you can defend.

Size your market →

Made by IntelliForge AI— a Hyderabad-based AI agent and workflow automation studio aligned with India’s Bharat AI Mission. Chronicle is the second product in a portfolio of nine shipped AI apps. Founder: Girish Hiremath— 14+ years across compliance, banking, pharma, telecom, and IoT, now formalising the AI side with an M.Tech in Data Science & AI at IIIT Dharwad.