Seed company-building version

Agent88 investor deck // First Round

We started with one customer truth: businesses do not need another chatbot. They need the work finished.

Agent88 is building from real workflow pain: one painful loop, one paid pilot, one proof artifact, one reusable module at a time.

View deckRound terms

Fit thesis

First Round fit: seed craft, PMF learning, founder insight, and company-building support. The emphasis is customer discovery, wedge discipline, and the learning loop from pilots to modules.

Presentation

Pre-seed investor deck.

Draft investor-facing narrative. External circulation should wait for Clyde/Yock approval, latest pilot status, and data-room readiness.

01

Problem

Companies bought AI tools, but work still falls through the cracks.

  • • Client follow-ups still depend on memory and manual discipline.
  • • Reports, proposals, research, and approvals are scattered across email, chat, Drive, PDFs, and SaaS dashboards.
  • • Copilots can answer questions, but businesses need governed execution across real workflows.

02

Why now

LLMs made knowledge work programmable. The next bottleneck is governed execution.

  • • Models can now read, reason, write, search, and operate tools.
  • • Teams are overwhelmed by disconnected AI copilots and fragile automations.
  • • Trust-sensitive workflows require human approvals, audit trails, and business memory.

03

Solution

Agent88 is a private agentic orchestration layer for enterprise workflows.

  • • Hermes maintains memory, synthesis, and institutional context.
  • • OpenClaw executes across browsers, APIs, files, documents, screenshots, and systems.
  • • Kanban Governance routes specialist agents, ownership, status, logs, and approval gates.
  • • The Human Board retains judgment, signoff, and accountability.

04

First wedge

Start with painful repeatable workflows in high-trust SMEs and agencies.

  • • Campaign Proof / Reporting Agent: creative assets + placement context + template → reviewed proof pack.
  • • Proposal / Document Agent: meetings + files + client context → proposal scope, quote draft, follow-up.
  • • Research Ingestion Agent: links/files/posts/PDFs → structured records, digests, routed tasks.
  • • Funding Navigator Plugin: company facts + project goal → funding screen, missing docs, draft support.

05

Proof path

RDS campaign proof automation is the first visible workflow wedge.

  • • Before: manual screenshot handling, asset placement, deck assembly, and client-ready formatting.
  • • Agent88 workflow: campaign asset + target placement context + report template → mockups → proof/report pack → human review.
  • • Why it matters: saves repetitive production time, improves client approval speed, and becomes reusable reporting automation IP.
  • • Caveat: RDS is a proof path; do not treat as booked revenue until signed/paid evidence exists.

06

Business model

Paid pilots → managed retainers → private deployments → reusable modules.

  • • Paid workflow pilots around one painful, bounded workflow.
  • • Managed agent retainers for monitoring, iteration, maintenance, and support.
  • • Enterprise/private deployments for client-scoped infrastructure, governance, integrations, and security posture.
  • • Module catalog: reporting, proposals, research, funding, CRM follow-up, and production coordination.

07

GTM

Sell outcomes first, productize the repetition.

  • • Founder-led sales through Hong Kong business and strategic finance relationships.
  • • Workflow audit CTA: identify one painful loop, scope a paid pilot, deliver proof, convert to retainer.
  • • Localised PR tracks: English/global, Hong Kong Traditional Chinese, Taiwan Traditional Chinese.
  • • Expansion from HK SMEs into Asia workflow buyers and enterprise innovation teams.

08

Moat

The compounding asset is workflow memory plus governed module reuse.

  • • Each pilot becomes a reusable execution module with prompts, tools, checks, acceptance criteria, and approval paths.
  • • Business-context memory and audit logs improve reliability across iterations.
  • • Private deployment and human approval make Agent88 suitable for trust-sensitive workflows.
  • • The system compounds from services-like wedge into infrastructure-like operating layer.

09

Roadmap

Twelve-month target: from proof to managed workflow infrastructure.

  • • 0–3 months: close 1–2 paid pilots, finalize RDS-style reporting module, publish proof/demo.
  • • 3–6 months: convert pilots into retainers, launch proposal/document and research modules, add client approval dashboard.
  • • 6–12 months: 2–3 managed retainers, private deployment package, Taiwan/global localisation tests, first physical-production coordination workflow.

10

Fundraise

Raising USD 400k pre-seed at USD 2.0M pre-money.

  • • USD 2.4M post-money; approximately 16.7% round ownership.
  • • Use of funds: product engineering, pilot delivery, customer success, governance dashboard, GTM, proof-pack/demo production.
  • • Milestone: prove that early workflow pilots can become repeatable modules and managed retainers.
USD 400kpre-seed raise
USD 2.0Mpre-money valuation
USD 2.4Mpost-money valuation
16.7%approx. round ownership