Book a call

03 / AI product · Workflow design · GTM operations

PitchKit AI

PitchKit converts rough client notes into campaign direction, platform-ready copy, production prompts, follow-up communication, CRM notes and a readiness score.

Role
Product strategy, workflow UX and the AI build
The gap
Conversation → brief. Context was rebuilt by hand.
The bet
Design the workflow, not the prompt.
Proof
Working product · Super9s Buildathon 2026 finalist
PitchKit AI010203040506
Role
Product strategy · UX / workflow design · AI product development
Product
AI workflow system for marketers and client-facing teams
Core capability
Guided intake → strategy → assets → QA → handoff
Principle
Human review stays inside the workflow

The handoff gap

Conversation →Brief

  1. 01ConversationRough notes + client context
  2. 02BriefAudience + offer + constraints
  3. 03StrategyAngle + rationale
  4. 04AssetsCopy + scripts + prompts
  5. 05QAReadiness + channel fit
  6. 06HandoffExport + activation

Where it brokeA productive client conversation held enough to act on, but the operator had to rebuild that context by hand for every brief, script and handoff.

The repairGuided intake carries the client's context forward, and every later step reads from it instead of starting again.

The repeated problem was translation.

Every client conversation produced enough information to act on. Then the operator rebuilt it, piece by piece, for each tool and each teammate downstream. The work was not slow because people were slow. It was slow because context did not travel.

Messy information → structured decision → usable work.

Decision log

  1. Obvious move
    Ship a better prompt template and let operators paste their notes in.
  2. Why not
    A prompt can generate content, but it drops channel constraints and operator judgment. Context would still leak at the same seam.
  3. The bet
    Make the decisions explicit: guided intake → strategy → assets → QA → handoff, with human review inside the flow.
  4. The tell
    The Readiness Score evaluates a brief before polish, so weak inputs fail early instead of producing confident-looking assets.
  5. Next
    With the first paying teams, measure output quality: how much of what PitchKit produces ships as is.

Design the workflow, not the prompt.

I defined the product, designed the workflow and built the AI. The product encodes the decisions that usually disappear between a client conversation and production.

Product strategy

I turned a repeated problem into a product: who uses it, what goes in, what comes out, and why that’s worth paying for.

Workflow + UX

I built intake, strategy, production, checks and handoff as one connected flow.

AI + operator knowledge

The AI carries client requirements forward, respects channel constraints and checks a brief is ready before it writes anything polished.

At the Buildathon, the feedback was that many people are trying to build this, and that automating a marketing team’s workflow is what they want.

Proof

  • 6Outputs from one set of rough notes
  • FinalistSuper9s Buildathon · Waterloo · May 2026
  • 10–15Teams in the Buildathon

I can find repeated operational friction, define the product logic and build a working AI workflow around human judgment.