Automate Proposal Generation & Project Intake for Consultancies with n8n + Claude Step-by-Step Guide

Intake Form Capture

{
  "source": "typeform" | "tally" | "hubspot_note" | "partner_initiated",
  "intake_id": "intake_01HXYZ...",
  "received_at": "2026-05-03T09:14:22Z",
  "signature": "sha256=...",
  "client": {
    "company_name": "Apex Regional Bank",
    "naics": "522110",
    "region": "EMEA / DACH",
    "headcount_band": "1000-5000",
    "revenue_band_usd": "500M-1B",
    "msa_in_place": false
  },
  "ask": {
    "problem_statement": "We need to evaluate our data strategy ahead of...",
    "desired_outcome": "A 12-month roadmap with prioritized initiatives and...",
    "deadline_driver": "board_meeting_2026-08-15",
    "phase_requested": ["assessment", "design"]
  },
  "commercial": {
    "budget_band_usd": "250k-500k",
    "procurement": "RFP",
    "prior_spend_with_firm_usd": 0
  },
  "discovery": {
    "call_recording_url": "https://gong.io/...",
    "transcript_text": "...",
    "stakeholders": [
      { "name": "Sarah Mueller", "title": "CDO", "role": "sponsor" },
      { "name": "Thomas Bauer", "title": "Head of Data", "role": "champion" }
    ]
  }
}

Project Classification

You classify a consulting intake into a strict taxonomy.
Read intake.ask and discovery.transcript_text only.
Never invent fields not present in the input.

PRACTICE AREAS (pick one or "blend"):
- strategy, operations, technology, finance_risk, people

PROJECT TYPES (pick one):
- assessment, target_operating_model, transformation,
  ma_diligence, regulatory_response, capability_audit, other

COMPLEXITY TIERS:
- T1: single practice, single region, <10 stakeholders, <90 days
- T2: blend, single region, 10-30 stakeholders, 90-180 days
- T3: blend, multi-region or regulated, 30+ stakeholders, 180+ days

DISQUALIFIERS to flag:
- conflict_of_interest_with_existing_client
- sanctioned_jurisdiction
- below_min_fee_50k
- scope_outside_practice (legal opinions, audit attest)

OUTPUT — strict JSON, no prose:
{
  "practice_area": "strategy" | "operations" | ...,
  "project_type": "assessment" | ...,
  "phase_mix": { "diagnose": 0.4, "design": 0.6, "build": 0, "run": 0 },
  "complexity_tier": "T1" | "T2" | "T3",
  "estimated_duration_weeks": int,
  "disqualifiers": [],
  "rationale": "2-3 sentences citing specific intake language"
}

Past-Project Retrieval

CREATE EXTENSION IF NOT EXISTS vector;

CREATE TABLE engagement_chunks (
  id              uuid PRIMARY KEY DEFAULT gen_random_uuid(),
  engagement_id   text NOT NULL,
  notion_page_id  text NOT NULL,
  chunk_text      text NOT NULL,
  chunk_kind      text NOT NULL,  -- abstract|sow|postmortem|winloss
  practice_area   text,
  project_type    text,
  complexity_tier text,
  region          text,
  signed_value_usd numeric,
  signed_at       date,
  embedding       vector(1536),
  updated_at      timestamptz DEFAULT now()
);

CREATE INDEX idx_chunks_embedding
  ON engagement_chunks
  USING hnsw (embedding vector_cosine_ops);

CREATE INDEX idx_chunks_filters
  ON engagement_chunks (practice_area, complexity_tier, region);

-- Pull the 8 most similar abstracts and 4 SOW excerpts
-- filtered by practice + complexity. The intake embedding
-- is computed once per proposal and reused.

WITH abstracts AS (
  SELECT engagement_id, chunk_text, signed_value_usd,
         signed_at, embedding <=> $1::vector AS dist
  FROM   engagement_chunks
  WHERE  chunk_kind = 'abstract'
    AND  practice_area = $2
    AND  complexity_tier = $3
  ORDER  BY embedding <=> $1::vector
  LIMIT  8
),
sow_excerpts AS (
  SELECT engagement_id, chunk_text,
         embedding <=> $1::vector AS dist
  FROM   engagement_chunks
  WHERE  chunk_kind = 'sow'
    AND  practice_area = $2
  ORDER  BY embedding <=> $1::vector
  LIMIT  4
)
SELECT * FROM abstracts
UNION ALL
SELECT engagement_id, chunk_text, NULL, NULL, dist FROM sow_excerpts;

Effort Estimation

// Inputs:
//   $json.grid           = phase × role hour matrix from Claude
//   $json.classification = { complexity_tier, project_type, ... }
//   $env.RATE_CARD       = role -> blended hourly rate USD

const RATES = JSON.parse($env.RATE_CARD);

// Buffer policy by complexity tier
const BUFFER = {
  T1: 0.12,   // 12% on simple, single-practice work
  T2: 0.18,   // 18% on blends or regional spans
  T3: 0.25    // 25% on regulated / multi-region T3s
}[$json.classification.complexity_tier];

// Risk add-on for known scope-creep triggers
const RISK_ADDONS = {
  data_migration: 0.05,
  regulatory_response: 0.07,
  multi_vendor_integration: 0.06
};

let baseHours = 0, baseDollars = 0;
for (const phase of Object.keys($json.grid)) {
  for (const role of Object.keys($json.grid[phase])) {
    const h = $json.grid[phase][role];
    baseHours   += h;
    baseDollars += h * RATES[role];
  }
}

const riskAddon = ($json.classification.risk_factors || [])
  .reduce((acc, k) => acc + (RISK_ADDONS[k] || 0), 0);

const totalBuffer = BUFFER + riskAddon;
const buffered = baseDollars * (1 + totalBuffer);
const fixedFee = Math.ceil(buffered / 1000) * 1000;

return [{ json: {
  base_hours: baseHours,
  base_dollars: Math.round(baseDollars),
  buffer_pct: Number(totalBuffer.toFixed(3)),
  fixed_fee_usd: fixedFee,
  effective_blended_rate: Math.round(fixedFee / baseHours)
}}];

Claude SOW Drafter

You are an SOW drafter for a mid-sized management consultancy.
Your output will be reviewed and edited by a partner before sending.
Your job is to land a 90%-finished draft, not a perfect one.

INPUTS PROVIDED:
- intake (problem statement, stakeholders, constraints)
- discovery transcript (verbatim, optional)
- 8 past-project abstracts ranked by similarity
- 4 SOW excerpts from comparable engagements
- effort grid (phases x roles x hours) and fixed_fee_usd
- firm voice guide (tone, banned terms, preferred section names)

WRITING RULES:
- Use only language the partner would use in front of the client
- No filler, no superlatives, no "synergy" or "leverage" as verbs
- Every claim about us must be defensible from past projects
- Every assumption must be falsifiable in the assumptions section
- Cite past engagements by anonymized type, never by client name

OUTPUT — strict JSON in this order:
{
  "situation": "1 paragraph, mirrors client's own framing",
  "objectives": ["3-5 bullets, outcome language"],
  "approach": "3-5 paragraphs, named methodology where applicable",
  "deliverables": [{ "name": "...", "description": "...", "phase": "..." }],
  "milestones": [{ "week": int, "label": "...", "exit_criteria": "..." }],
  "team": [{ "role": "...", "allocation_pct": int, "rationale": "..." }],
  "fees": {
    "structure": "fixed" | "tm" | "phased",
    "amount_usd": int,
    "payment_schedule": [...]
  },
  "assumptions": ["..."],
  "exclusions": ["..."],
  "risks_for_partner_review": ["flagged uncertainties"]
}

NEVER output: fabricated client logos, made-up case studies,
unverified statistics, or any past client name.

{
  "method": "POST",
  "url": "https://api.anthropic.com/v1/messages",
  "headers": {
    "x-api-key": "{{ $credentials.anthropic.apiKey }}",
    "anthropic-version": "2023-06-01",
    "content-type": "application/json"
  },
  "body": {
    "model": "claude-sonnet-4-5",
    "max_tokens": 6000,
    "system": "{{ $json.sowSystemPrompt }}",
    "messages": [
      {
        "role": "user",
        "content": [
          { "type": "text",
            "text": "Intake:n{{ JSON.stringify($json.intake) }}" },
          { "type": "text",
            "text": "Classification:n{{ JSON.stringify($json.classification) }}" },
          { "type": "text",
            "text": "Past-project context:n{{ $json.retrievalText }}" },
          { "type": "text",
            "text": "Effort grid:n{{ JSON.stringify($json.grid) }}nFixed fee: ${{ $json.fixed_fee_usd }}" },
          { "type": "text",
            "text": "Discovery transcript (verbatim):n{{ $json.intake.discovery.transcript_text }}" }
        ]
      }
    ]
  }
}

Partner Review Queue

{
  "channel": "#proposal-review",
  "blocks": [
    {
      "type": "header",
      "text": { "type": "plain_text",
                "text": ":memo: T2 SOW draft — Apex Regional Bank" }
    },
    {
      "type": "section",
      "fields": [
        { "type": "mrkdwn",
          "text": "*Practice*nStrategy + Data" },
        { "type": "mrkdwn",
          "text": "*Fee*n$278,000 fixed" },
        { "type": "mrkdwn",
          "text": "*Duration*n14 weeks (Diagnose + Design)" },
        { "type": "mrkdwn",
          "text": "*Closest comparable*nEU regional retail bank, 2025-Q3, won at $312k" }
      ]
    },
    {
      "type": "section",
      "text": { "type": "mrkdwn",
                "text": "*Flagged for partner review:* T&M vs fixed structure (data migration risk), CDO requested by name in transcript but is sponsor not champion." }
    },
    {
      "type": "actions",
      "elements": [
        { "type": "button",
          "text": { "type": "plain_text", "text": "Open in Google Docs" },
          "url": "https://docs.google.com/document/d/{{docId}}" },
        { "type": "button",
          "text": { "type": "plain_text", "text": "Approve & send" },
          "style": "primary", "action_id": "approve_sow" },
        { "type": "button",
          "text": { "type": "plain_text", "text": "Re-draft with notes" },
          "action_id": "redraft_sow" }
      ]
    }
  ]
}

DocuSign + CRM Sync & Kickoff

{
  "method": "POST",
  "url": "https://eu.docusign.net/restapi/v2.1/accounts/{{accountId}}/envelopes",
  "headers": {
    "authorization": "Bearer {{ $credentials.docusign.access_token }}",
    "content-type": "application/json"
  },
  "body": {
    "templateId": "{{ $env.SOW_TEMPLATE_ID }}",
    "templateRoles": [
      {
        "email": "{{ $json.intake.discovery.stakeholders[0].email }}",
        "name":  "{{ $json.intake.discovery.stakeholders[0].name }}",
        "roleName": "Client Signer",
        "tabs": {
          "textTabs": [
            { "tabLabel": "ProjectName",
              "value": "{{ $json.sow.project_name }}" },
            { "tabLabel": "FixedFee",
              "value": "${{ $json.fixed_fee_usd }}" },
            { "tabLabel": "DurationWeeks",
              "value": "{{ $json.sow.duration_weeks }}" }
          ]
        }
      },
      {
        "email": "{{ $json.partner.email }}",
        "name":  "{{ $json.partner.name }}",
        "roleName": "Firm Signer"
      }
    ],
    "status": "sent",
    "emailSubject": "{{ $json.firm.name }} — SOW for {{ $json.intake.client.company_name }}",
    "customFields": {
      "textCustomFields": [
        { "name": "deal_id_hubspot", "value": "{{ $json.deal_id }}" },
        { "name": "intake_id",       "value": "{{ $json.intake_id }}" }
      ]
    }
  }
}

{
  "method": "PATCH",
  "url": "https://api.hubapi.com/crm/v3/objects/deals/{{ $json.deal_id }}",
  "headers": {
    "authorization": "Bearer {{ $credentials.hubspot.token }}",
    "content-type": "application/json"
  },
  "body": {
    "properties": {
      "dealstage":            "proposal_sent",
      "amount":               "{{ $json.fixed_fee_usd }}",
      "closedate":            "{{ $json.sow.expected_close_date }}",
      "skru_proposal_id":     "{{ $json.proposal_id }}",
      "skru_complexity_tier": "{{ $json.classification.complexity_tier }}",
      "skru_practice_area":   "{{ $json.classification.practice_area }}",
      "skru_buffer_pct":      "{{ $json.buffer_pct }}",
      "skru_docusign_env_id": "{{ $json.envelope_id }}"
    }
  }
}