Credibility — Two years of applied AI for clients

What this slide says

Two years of hands-on applied AI work directly with clients — limited-scope on any given engagement, but real, shipped, and grounded in actual client needs. The work spans fundraising, dealflow, memo/deck automation, website rebuilds, content marketing, and market research. Underneath the visible outputs, three load-bearing areas of focus are consistent across engagements: toolkit selection and adoption, RAG corpus generation and connectors, and realtime data integrations. The outcomes show up where clients live — better fundraise efficacy, less dealflow anxiety, actionable portfolio monitoring, less single-vendor lock-in, and broader AI/LLM use to lift output while lowering stress.

Why this slide exists

The credibility anchor for the deck. Investors / clients / partners reading this need to see “this isn’t theory — we’ve been shipping with clients for years.” Job in the arc: convert curiosity to trust early enough that everything downstream (vision, capability, ask) has something to stand on. Without this slide the rest of the deck reads as aspirational rather than reportable.

Content

The claim

Two years of applied AI work, directly with clients, on real (but often limited-scope) needs.

Outputs (what we shipped)

• Fundraising systems
• Dealflow systems
• Memo and deck draft automations
• Website rebuild or augmentation
• Content marketing systems
• Market research systems

Three load-bearing focus areas (how we shipped them)

1. Toolkit selection and adoption — picking the right tools for a given client and getting them actually used inside the team’s workflow.
2. RAG corpus generation, curation, and filesystem connectors — turning a client’s scattered source material into a retrievable, queryable substrate.
3. Realtime data source integrations — wiring live data into the surfaces where decisions get made.

Outcomes (what changed for the client)

• Fundraising: pipeline efficacy and pitch efficacy improvements
• Dealflow: improved dealflow awareness; decreased anxiety and workload on dealflow processing
• Portfolio: monitoring improvements with actionable interventions
• Vendor posture: decreased reliance on single or proprietary systems
• Team posture: increased use of AI/LLM to lift output and reduce stress

Layout candidates

• 3-column “how” layout — toolkit / RAG / realtime as equal pillars; outputs in a strip above, outcomes in a strip below
• Outputs-then-outcomes vertical — outputs left column, outcomes right column, the three focus areas as connective tissue between them (arrows or a center spine)
• Single-row evidence strip — small client logos under each output category (the “we shipped this for this client” proof move)

Substantiation

• Calmstorm engagement — fundraising materials, dealflow surfaces (client-sites/calmstorm-decks)
• Chroma engagement — pitch deck + investor-roster credibility cards (client-sites/chroma-decks)
• Humain engagement — operating-company deck (client-sites/humain-vc-decks)
• Reach Education engagement — education-sector precedent (client-sites/reach-edu-hub)
• Podium Education — imminent engagement, surfaced in context-v/explorations/Cloud-Variant-of-Dididecks-AI-Workspace.md
• The dididecks-ai project itself — meta-evidence; the deck infrastructure was authored by doing this work
• ai-labs/memopop-ai — investment-memo orchestration as a sibling capability

Client-specific outcome metrics (efficacy lifts, anxiety reductions) want sourcing — at minimum a quote per client, at best a measurable. Open question in Authoring notes below.

Authoring notes

• Outputs vs. outcomes tension. They look adjacent but do different work: outputs are what we built; outcomes are what changed for the client. Keep the distinction sharp on the slide — if a reader has to squint to tell them apart, the slide is over-packed.
• “Limited-scope” qualifier. Honest but could dilute credibility on a skim. Recommendation: keep it in the body / speaker notes for one-on-one conversations, drop it from the headline. The qualifier is what makes the claim true but it’s not what a partner reads off a screen in 8 seconds.
• The three focus areas. They’re the strongest part of the slide because they show what’s repeatable, not just what was done. Worth visual weight — center column, larger type, or a numbered spine.
• Client evidence. Every output should be backed by at least one client logo or named engagement. If a category has no client behind it, downgrade the claim or drop the category for this deck.
• Outcome metrics. “Decreased anxiety” is a real thing clients say but doesn’t quantify. For the DD-grade version of this slide, every outcome wants either a number or a sourced quote. For a teaser version, the qualitative claims survive.
• Order in the deck arc. Likely follows the vision/hero and precedes the capability/methodology slides — credibility before specifics. Reorder is cheap; this is just the default.

Open questions

• Does this slide hold its own as one slide, or want a two-slide split (Outputs / Outcomes) for breathing room?
• Should the three focus areas get a dedicated slide of their own elsewhere in the deck, with this one referring back?
• Is there a version that leads with one engagement’s story (Calmstorm as exemplar) and uses the others as a quieter strip below, instead of treating all engagements as equal?