CONSELA

The decision process you would design
if you had unlimited budget.

A perfect advisory team: AI advisors who show up prepared, cite their sources, and challenge poor reasoning. They research your question in real time and debate it into a decision grounded in verified research.

14-day free trial · 100 credits · no card required · Powered by OpenAI GPT-6

Independent AI advisors Real-time research Every citation verified 30 languages 14-day free trial, no card Zero-retention inference

The problem

Just because you can decide alone doesn't mean you should.

A single perspective can't see its own blind spots — no matter how capable the tool behind it. So CONSELA runs separate investigations, an advisor whose job is to disagree, and a debate you can read line by line — ending in a decision grounded in verified research.

One point of view

  • One lens on your question
  • Nothing pushes back on the answer
  • Claims you can't easily verify

Independent perspectives

  • Several lenses, researched separately
  • Every claim is challenged before it stands
  • Verified sources — references that can't be traced are stripped

How it works

One decision. Five stages.

The difference isn't fancier language — it's the structure: independent research, an advisor who disagrees, a debate you can read line by line, a consensus document you can audit, and a follow-up loop that can produce a new consensus when your questions change the picture.

  1. 01

    Research

    AI advisors investigate your market in parallel, each citing live, up-to-date sources.

  2. 02

    Challenge

    One advisor is hired to disagree — it reads every report and re-researches them, so the blind spots surface before you commit.

  3. 03

    Debate

    The AI advisors test every recommendation against each other's evidence — in writing, with an attributed transcript.

  4. 04

    Consensus

    A consensus document focused on outcomes: decisions, owners, action items, and a risk register — every citation verified.

  5. 05

    Follow-up

    Challenge an assumption or ask for more depth. Your follow-up can update the analysis and generate a new consensus.

The difference

Every citation is verified.

A source only appears if it was verified by advisor research or safely checked from your meeting brief. What can't be traced never reaches your reports.

researchgate.net/publication/3188736 — cited, never visited Stripped before you see it
europa.eu/publications/sme-finance — visited and checked Rendered as a verified source

The advisory team

10 AI advisors. One decision.

Choose the perspectives that fit the decision. Selected advisors research on their own, then have to hold their ground in front of the others — with the Skeptic always included to challenge the findings.

Finance

Financial analysis, runway, pricing

Operations

Process, scaling, efficiency

Sales & Marketing

Pipeline, GTM, customer acquisition

Technology

Architecture, engineering, tech debt

Strategy

Positioning, competitive moat, growth

Risk

Risk assessment, compliance, threats

Lawyer

Your legal risk advisor for contracts, obligations, and regulatory questions

Competitors & Industry Specialist

The senior specialist on your field, that knows and researches everything about your competitors

Customer Satisfaction

Your specialist in customer satisfaction, it puts itself on the customer shoes

Skeptic

Cross-examines all reports for blind spots

Quiet by design

No noise. Just the decision.

One structured flow on a calm interface — the work gets the screen, not the chrome.

One structured flow

Your question, five stages, one consensus document. Choose which advisors participate in each meeting; the required Skeptic stays in the room, and selecting fewer advisors uses fewer credits.

A calm interface

No widget walls, no metric noise — the screen does one thing: shows the research as it happens, without distraction.

Output you can read

An attributed transcript, verified sources, and a consensus document built for action — not an endless chat log.

After the meeting

One decision is never the last.

A decision is rarely one conversation. Your project keeps the record, so you can bring relevant context into future meetings.

Your document library

Upload once and it stays in your project. Project files and extracted text are encrypted at rest, while meetings and follow-up answers stay grounded in your own material.

Project memory

Completed conclusions stay with your project. Bring relevant previous decisions into a new meeting when they matter.

Follow-up questions

Challenge any report, ask for more depth, or push back on a decision — the debate continues.

Share the record

Every meeting ends with a shareable link — send the consensus document and its verified sources to a co-founder or investor.

What founders use it for

Decisions that deserve more than a quick answer.

  • Pricing a product

    Runway, willingness-to-pay, tier design — argued and documented before you charge.

  • Entering a market

    Size the opportunity, map the competitors, and stress-test the entry strategy.

  • Build vs. buy

    Technology, cost and risk perspectives on the call that's hard to reverse.

  • Raising or budgeting

    Realistic finances from your actual numbers — pricing, runway, and a budget that holds up when investors ask.

  • Regulation & compliance

    New rules landing across borders? Map what changes, what it costs, and your compliance plan before it becomes a fine.

  • Hiring & people

    Spec the role, benchmark real pay in your market, and decide hire vs. outsource before you post it.

Built for sensitive decisions

Your organization’s information stays under your control.

CONSELA keeps your meetings, documents, answers, and follow-ups inside your organization’s workspace. Project access controls determine which members can see a decision, and encryption protects the information while it is stored and transferred.

Organization-scoped access

Your project members see only the projects and meetings they have been given access to. They are separate from the AI advisors.

Protected in transit and storage

Your workspace content is protected by access controls and encryption while it moves through the app and while it is stored.

Controlled AI processing

Relevant content is sent only to configured AI providers needed for your requested analysis, using verified zero-data-retention routes.

No general model training

CONSELA does not use your organization’s information to train general-purpose models. Web research providers receive only the data needed for the requested search.

You decide what to submit and who can access each project. Read the Privacy Policy View Subprocessors

Questions

Asked before the first meeting.

How is this different from asking ChatGPT?

Ask one tool — ChatGPT, Claude, a consultant — and you get a single point of view with no one to disagree with it. CONSELA runs separate investigations, an advisor whose job is to disagree with every finding, then a written debate. The result is a decision grounded in verified research — not a single answer.

How is this different from other AI advisory tools?

CONSELA does more than provide an open chat or a static advisory workspace. Its advisors research your decision live across relevant web sources during the meeting, cite verified sources, challenge the findings, debate the evidence, and produce one readable consensus document you can question afterwards.

Which AI models power CONSELA?

CONSELA currently uses OpenAI GPT-6 as its default AI model. The app administrator may update the model as providers, pricing, availability, and safety requirements change.

Can I choose which advisors work on my decision?

Yes. Select the advisors whose perspectives fit the question when you start a meeting. The Skeptic is always included, and choosing fewer advisors can reduce credit usage.

What kinds of decisions can CONSELA help with?

Business decisions such as pricing, market entry, hiring, supplier selection, budgeting, partnerships, compliance, and product priorities. Bring a specific question and the context you have; the exact analysis depends on your brief and evidence.

How does advisor selection affect credits?

Credits cover the meeting workflow, including advisor research and model usage. Selecting fewer advisors generally means less advisor research and lower usage, while the actual amount depends on the work performed and research required.

Can I use my own documents and previous meetings?

Yes. Add project documents and optionally load completed meetings from the same project so the selected advisors can use that context while re-checking what earlier work found.

Can I compare options rather than get a single recommendation?

Yes. Frame the alternatives and the criteria that matter in your brief. The advisors can research the trade-offs and produce a documented comparison with a recommendation and risks.

Can I challenge the result with follow-up questions?

Yes. Ask about a passage, challenge an assumption, request more depth, or address a specific advisor. Follow-up threads stay connected to the decision and can be applied to a new consensus version.

Can my team access and receive updates about a decision?

Yes. Project members are people from your organization who have access to the project. You can choose which members receive in-app updates when a meeting needs input or finishes. They are separate from the AI advisors selected to research the decision.

What should I include in the meeting brief?

Describe the decision, your market and customers, relevant constraints, alternatives, and what a useful recommendation would help you decide. Add documents or previous meeting context when they contain evidence the advisors should examine.

Are the sources real?

Yes — and verifiable. A URL only appears as a source if advisor research verified it or it was safely checked from your brief. References that can't be traced are stripped before you see the reports, and every cited claim points back to a source you can open yourself.

How fresh is the research?

AI advisors research live during your meeting, and the evidence gate requires at least two sources verified through advisor tool activity. If the market shifted this week, the research reflects it.

How long does a meeting take?

Total wait time depends on the queue, the number of advisors, and the complexity of your meeting. Most meetings finish within 2–10 minutes once processing starts, but larger or more complex meetings can take up to 30 minutes. Meetings run in the background, so you can do something else while the advisory team works. When it finishes, you receive an in-app notification and an email. If the AI advisors need more information, you receive the same notifications with a request to answer their questions.

What happens to my data?

Your meetings, documents, answers, and follow-ups stay within your organization’s workspace and are protected by access controls and encryption. We process them to provide the research, AI analysis, documents, notifications, and follow-up features you request; relevant content may be sent to AI providers for inference using configured zero-data-retention routes. Web-research providers receive only the data needed for the requested search and process it under their applicable terms. These requests are not used to train general-purpose models by CONSELA. See the Privacy Policy and Subprocessor List for provider and international-processing details.

Is CONSELA suitable for every decision?

No. CONSELA provides AI-generated advisory material, not professional advice, and should not be the sole basis for decisions about employment, credit, housing, education, healthcare, insurance, or essential services. Review important information with qualified professionals.

Which languages can I use?

30 languages, from English, Spanish and Portuguese to Japanese, Arabic and Hindi. AI advisors meet and answer in your language.

What does it cost?

Flat monthly plans from $20 — one shared monthly credit allowance with usage shown after each action. Start free: 100 credits for 14 days, no card required. See pricing.

Daniel Rossi Marinho, founder of CONSELA

Daniel Rossi Marinho

Founder, CONSELA · Portugal

The person behind CONSELA

Built by someone who has spent a decade making business software work.

Running a business often means making consequential decisions with limited time, scattered information, and nobody available to challenge the first idea.

I built CONSELA to make that process more rigorous. I bring more than 10 years of experience building software systems, automation, integrations, data platforms, and infrastructure to a product that researches the question, examines it from different perspectives, and challenges the findings.

The goal is not to replace your judgment. It is to give you better evidence, stronger opposition, and a record of what was decided and why.

Your decision is one question away.

Start free — 100 credits, no card required.