ThoughtProcess

Rigorous AI

For work that demands specifity and project-specific institutional knowledge, generic AI does not have the full context it needs. Thought Process closes the gap by building a structured knowledge base, reflecting how your team actually thinks about its materials. AI then helps navigate and augment your own analysis, rather than flattening it or offering generic input.

Rich, team-specific context for better AI systems

  • Your analysis first
    Start with your own sources and team's commentary. Everything is structured automatically to be used as context for AI sessions: ask the model to recollect and synthesize several threads at once, distill disparate analyses, and augment complex thinking.
  • Wherever you prefer
    Use built-in chat, or bring your own: use the tools provided by our MCP server using the model of your choice, through any MCP client.
  • Your data is yours
    Your material is never used to train models. When using the MCP server, we just provide the tools, not models, so your AI usage operates under the data agreement you already have.

Don't settle for hallucinations

Our AI focuses on detailed, accurate responses based on your own sources and your team's thinking about them. Every response cites specific resources from your knowledge base: passages from sources, notes, team discussion, writing, and past searches.

Prioritize your perspective

Start from what you deem important in your own material. Augment your team's unique analysis using AI that is equipped with the precise context it needs.

We go beyond generic summaries: our tools are built to focus on the specifics of your documents and how they feature into your work.

Fully grounded and traceable AI responses, oriented around the structure you've put on your knowledge base.

Connect any AI client to our MCP server. Access your library and our toolkit from coding agents, RAG pipelines, or other systems you already use.

Build on your own thinking

Drop in resources from the workspace as you go. Conduct in-depth synthesis and analysis on the fly: summarize full collaborator discussion, compare to your past notes on a given subject, find everything you have on a nuanced topic.

AI for precise, nuanced analysis.

←Structure
Collaboration→
ThoughtProcess
Product
  • Collections
  • Search
  • Structure
  • AI & MCP
  • Writing
  • Collaboration
Resources
  • Security
  • Deployment
  • Getting started
Company
  • Pricing
  • Contact
  • Privacy & Terms
© Thought Process Software, LLC.info@thoughtprocess.us
ACME Inc. - FY26 review
AISearchSources
Upload
Looking for...

Sources

New folder
  • 10-K ACME Inc.Details
    JJakeJake· added March 4, 2026
    1 notein 2 documents1 threadin 2 searches
    You · March 16, 2026
    1

    Margin bridge doesn’t reconcile with the segment tables. Pull the Q2 10-Q before this goes out.

    Linked sections
  • ACME Inc. 10-Q - Q2 2026
  • ACME Inc. earnings call - Q2 2026 transcript
  • Margin analysis
    • Ridgeline Research - ACME downgrade to hold
    • Meridian Partners - ACME channel checks, Q1
    • Cost basket
      • BLS producer price index - industrial inputs
      • Segment margin bridge - working modelDetails
        JJohnJohn· added June 9, 2026
        2 notesin 1 document1 threadin 3 searches
NotesSearch

Ridgeline Research - ACME downgrade to hold

We are moving to Hold. The Q3 print did not settle the question the Q2 print raised, and management’s framing of the margin bridge has not improved.

Our basket of ACME’s principal industrial inputs - resins, fabricated steel, and inbound freight - rose 6.1% year over year against a cost of goods base near 62% of revenue, implying 80 to 95 basis points of gross margin pressure. Attribution of the full 240 basis points to input costs is therefore incomplete.

We would revisit on segment-level price and volume disclosure, which the company has declined to provide for three consecutive quarters.

Without it the bridge cannot be rebuilt: our basket leaves 145 to 160 of the 240 basis points unexplained.