The 5 AI Experts That Review Every Deck Before You Send

Five AI experts. Every deck. Every page. Before it leaves your firm.

That is the baseline for every proposal reviewed by Lurio — five curated experts active out of the box, trained to catch what the creator cannot see. This post introduces each one: what they look for, what they typically catch, and why it matters for the agencies and consulting firms that rely on proposals to win work.

These are the five curated experts — the first tier of a three-tier system (curated → recommended → custom). They work for any organisation, on any proposal type, without configuration.

1. Strategy Critic

What they look for: Narrative logic, argument structure, and whether the thesis is supported by the evidence on the page. Not style — substance.

What they typically flag: Market size claims that lack a methodology. Conclusions that do not follow from the data shown. A solution presented before the problem has been established. An executive summary that promises proof the body of the deck does not deliver.

Example critique: "The market size claim on slide 4 is not supported by the figures cited. The TAM calculation should appear before the growth projection, not after. Currently the deck asks the reader to accept the conclusion before seeing the evidence."

Why it matters for agencies: Senior strategists are the bottleneck. Every agency has one — maybe two — partners who can read a proposal and tell you whether the argument actually works. The Strategy Critic does the first-pass review so the partner's time is spent on judgment and client context, not catching structural errors that should have been caught before it landed in their inbox.

2. Brand Compliance

What they look for: Visual and voice consistency against your brand guide — colours, fonts, tone, logo use. Every slide should be unmistakably on brand.

What they typically flag: Colour values that deviate from brand primaries. Font weights applied incorrectly. Logo misplacement or resizing outside approved bounds. Tone that shifts between slides — formal in the executive summary, casual in the appendix.

Example critique: "Slide 8 uses a secondary colour in a primary-action position. The section header on slide 12 uses a font weight not specified in the brand guide. These inconsistencies undermine the precision that the visual system is designed to communicate."

Why it matters for agencies: One off-brand slide in a client-facing proposal signals to the client that the agency does not pay attention to detail. That is exactly the opposite of what a proposal is supposed to demonstrate.

3. Narrative Reviewer

What they look for: Story arc, clarity, and persuasion. Does the deck take the reader from problem to solution to proof in the right order? Does each slide earn its place in the sequence?

What they typically flag: The client problem buried until slide 5. Slides that repeat the previous slide in different words. A conclusion that adds no new information. Transitions that break the logic flow — where the reader has to work to understand why this slide follows the last one.

Example critique: "The client problem is established on slide 5, but the solution is introduced on slide 3. The reader is asked to evaluate an answer before they understand the question. Move the problem statement to the opening section — before the solution, before the credentials."

Why it matters for agencies: The best-designed slide in the world fails if the narrative logic does not hold. Proposals win on argument, not aesthetics. The Narrative Reviewer ensures the argument is built correctly before the deck reaches the client.

4. Data Integrity

What they look for: Numerical consistency, source citation, and internal contradictions. The expert cross-checks every figure in the deck against every other figure and against your cited sources.

What they typically flag: Two figures that purport to measure the same thing but produce different results on different slides. A percentage that does not add up to 100 with its stated components. A market size figure that contradicts the growth projection. A benchmark that predates the period the deck is discussing.

Example critique: "The customer acquisition cost on slide 6 is stated as £1,200. The attribution model on slide 11 implies a CAC of £980 using the same period and channel mix. Reconcile before this reaches a CFO who will check both slides."

Why it matters for agencies: One data error in front of a data-literate prospect ends the conversation. The Data Integrity expert is the last line of defence before the numbers leave your firm.

5. Audience Fit

What they look for: ICP alignment. Are the proof points, case studies, and benchmarks calibrated for the specific audience receiving this proposal — their industry, their seniority, their decision criteria, their known objections?

What they typically flag: Sector-mismatched benchmarks. Case studies from the wrong vertical. Language calibrated for the wrong seniority level. Proof points that answer questions the prospect did not ask, while missing the ones they will.

Example critique: "This deck was configured for a fintech ICP. Slide 9 references retail-sector customer acquisition benchmarks. A fintech CMO will immediately identify these as irrelevant. Replace with the fintech benchmarks from your 2025 sector report."

Why it matters for agencies: A generic proposal is worse than no proposal — it signals that the agency treated this client like every other client. The Audience Fit expert ensures every proposal is calibrated for the specific reader before it leaves the building.

Beyond the Five: Recommended and Custom Experts

The five curated experts are the baseline. For firms with specific knowledge bases and client portfolios, Lurio generates recommended experts automatically — trained on your brand voice, your ICP definition, your past-winning work. You review, edit, or disable each one.

For specialist workflows, you build custom experts — a per-client Audience Fit expert trained on that client's documents, a CFO sign-off expert for a manufacturing client, a legal tone reviewer for RFP responses. No code required.

The five described above are the ones that run on every proposal, for every team, from day one. They are where the review starts.