deck-beamer-proposal-report

Beamer Proposal / Report Deck

Produces compile-ready Beamer source for proposals (what you intend to do and why it's fundable/approvable) and reports (what you did versus what you promised). These genres are structured around a plan-of-work, not a single contribution — they need objectives, work packages, timelines, milestones, deliverables, budget, and risk, and reports add progress-vs-plan.

This is the render step. Its slide-spec schema and format-independent design rules live in references/slide-spec-and-design.md — read that first. It is not deck-beamer-academic (single-paper results talk) and not deck-business-report (PPTX/Canva, executive audience).

Two genres, different spines

Proposal deck

Audience: committee / funder / advisor deciding whether to approve or fund. They evaluate significance, feasibility, and the credibility of the plan.

Spine:

Problem & significance — why this matters, why now.
Gap / state of the art — what's missing; what you'll do that others haven't.
Objectives / research questions — 2–4 crisp, testable aims.
Approach / methodology — per objective, how you'll attack it.
Preliminary results — evidence you can pull this off (de-risks the ask).
Work plan & timeline — work packages mapped to a Gantt.
Milestones & deliverables — concrete, dated, verifiable.
Risks & mitigations — name the real ones; show you've thought ahead.
Budget / resources (if applicable) — what you need and why.
Expected impact & summary — what success unlocks.

Report deck (progress / milestone / final)

Audience: same stakeholders, now checking delivery against the promise.

Spine:

Recap of objectives — what was promised (one slide).
Progress vs. plan — status per objective/work package (on track / at risk / done).
Key results so far — the evidence, claim-titled.
Milestones hit / missed — honest status table.
Deviations & why — changes from the plan and their justification.
Risks (updated) — what materialized, what's new.
Remaining work & revised timeline — what's left, by when.
Budget/resource status (if applicable) — burn vs. plan.
Summary & ask — decision or support you need next.

Workflow

1. Gather inputs

Genre (proposal / progress / mid-term / final).
The plan-of-record: objectives, work packages, original timeline, milestones, deliverables, budget.
For reports: current status per item, what slipped and why.
Audience and slot length; institutional/funder Beamer template if mandated.

2. Build the slide-spec

Use the relevant spine above as the skeleton, then produce per-slide spec blocks (references/slide-spec-and-design.md). For the persuasion craft behind a proposal deck — framing it as a sell not an info dump, the hourglass, the three proposal figure archetypes (butterfly / family-portrait / before-after), and density discipline — see references/proposal-persuasion-craft.md (grounded in Knapen et al., 2025, arXiv:2504.01645). Genre-specific guidance:

Objectives get claim-titles framed as outcomes, not topics.
Timeline → one Gantt slide (TikZ pgfgantt), not a wall of dates.
Milestones/deliverables → a compact status table (booktabs), ≤ 6 rows visible; overflow to backup.
Risks → a 2-column "risk → mitigation" (or likelihood/impact) table, top risks only.
Progress (reports) → a status table with a clear legend (✓ done / ◐ on track / ✗ at risk), color-coded but never color-only.
Budget → summarize categories; full breakdown goes to backup.

Confirm the spec before emitting LaTeX.

3. Emit compile-ready .tex

Render the approved spec using assets/preamble.tex (adds pgfgantt, status macros, and a risk/milestone table style on top of the academic preamble). Requirements:

Self-contained, compiles with pdflatex/lualatex; placeholders clearly % TODO-marked.
Gantt chart built with pgfgantt; milestone/deliverable/risk tables with booktabs.
Status macros (\done, \ontrack, \atrisk) that pair a symbol with a color (colorblind-safe).
Speaker notes via \note{}; section recap via \tableofcontents[currentsection] for longer reviews.
A backup-slides appendix for detailed budget, full milestone list, and method depth.

4. Hand off

Give the compile command, the files to drop in, where to edit dates/numbers, and which detail slides live in the appendix.

Ship scripts/compile.sh with the .tex so it builds in one step (latexmk if present, else two pdflatex passes; pgfgantt must be installed):

./scripts/compile.sh proposal.tex        # build
./scripts/compile.sh -c proposal.tex     # build, then clean aux files

What separates good proposal/report decks from paper talks

The plan is the product. A vague timeline or hand-wavy milestones sink a proposal faster than a weak result. Make every milestone dated and verifiable.
Feasibility beats novelty here. Preliminary results exist to show you can execute, not to be a full contribution.
Reports must be honest about slippage. A status table that hides "at risk" items destroys trust; show them with the mitigation.
Numbers are checkable. Budget, timelines, and milestone dates will be scrutinized — keep them consistent across slides.

Anti-patterns to refuse

Objectives written as topics ("Study X") instead of testable aims with success criteria.
A timeline slide that's a paragraph of dates instead of a Gantt.
A report that buries missed milestones or omits the progress-vs-plan view.
Cramming the full budget/work-package detail onto main slides instead of backup.

Output

The confirmed slide-spec (markdown), built on the proposal or report spine.
A single compile-ready .tex (plus preamble) with Gantt, status tables, and marked placeholders.
Compile + edit + appendix instructions.