Skip to content

skyplon/Prospect-IQ

Repository files navigation

ProspectIQ — AI Sales Intelligence & Outreach Generator

A full-featured enterprise sales tool built for SDRs who need to move from research to personalized outreach fast. ProspectIQ aggregates multi-source prospect signals into a unified intelligence layer and generates brand-compliant outreach emails — with a conversion learning loop that gets smarter with every email sent.


The Problem

Enterprise SDRs spend 60% of their working day manually researching prospects across financial reports, LinkedIn profiles, and tech stack databases before writing a single personalized email.

Despite that effort, conversion rates stall — because no human can simultaneously synthesize financial signals, behavioral signals, and tech context into a strategically relevant message at the speed and volume enterprise outreach requires.

Two compounding failures drive this:

  • Efficiency failure — data retrieval and synthesis is machine work being done by a human
  • Quality failure — manual synthesis produces shallow personalization that misses product-market fit

ProspectIQ solves both by replacing the 5-tab research loop with a single AI-powered intelligence and generation engine — with live data connections and a self-improving conversion model built in from day one.


Solution Overview

ProspectIQ is a standalone web-based Applied AI tool that gives Enterprise SDRs:

  1. A unified research surface — one prospect input triggers simultaneous intelligence gathering across three live signal streams
  2. A signal-to-message bridge — AI synthesizes multi-source signals into a coherent account narrative and personalized email
  3. Brand compliance by design — guidelines enforced at generation time, not as a manual afterthought
  4. Live data source integrations — real-time connections to LinkedIn Sales Navigator, S&P Capital IQ, BuiltWith, and Factiva — every signal current, timestamped, and confidence-scored
  5. A self-improving conversion loop — the model learns which signal combinations and message angles drive the highest reply and meeting rates, and weights future generation accordingly

Core Features

Live Intelligence Status Bar

A persistent dark bar below the header shows real-time connection status for every data source:

  • LinkedIn Sales Navigator — live · 82ms
  • S&P Capital IQ — live · 134ms
  • BuiltWith Intelligence — live · 61ms
  • Factiva News Feed — syncing

Each source displays a pulsing live indicator, an individual refresh button, a global signal record count (e.g. 2,457 signals), and a last-refresh timestamp — anchoring the entire intelligence feed to a real-time state the SDR can trust.

Intelligence Engine (Left Panel)

  • Prospect Input Form — Name, company, role, and outreach goal as the single starting point
  • Financial Signals Card — Recent earnings highlights, growth trajectory, headcount changes, and strategic priorities sourced via S&P Capital IQ. Each signal shows its source label (e.g. "via S&P Capital IQ"), a last-refreshed timestamp with a live pulse dot, and a per-signal confidence bar
  • Behavioral Signals Card — Recent LinkedIn posts and activity, engagement topics, role tenure, and shared connections sourced via LinkedIn Sales Navigator — live and continuously updated
  • Tech Stack Signals Card — Current tools, identified gaps, displacement opportunities, and integration fit sourced via BuiltWith Intelligence and Factiva News Feed
  • Synthesized Intelligence Summary — One AI-generated paragraph connecting all three live signal streams into a coherent account narrative

Email Output Engine (Center Panel)

  • Personalized Email Generator — Complete outreach email with hook, value bridge, and CTA generated from live intelligence signals
  • Three Variant Tabs with Live Reply Rates — Each tab displays its current conversion performance so the SDR instantly knows which angle is working:
    • LinkedIn Angle: 41% reply rate
    • Financial Angle: 28% reply rate
    • Tech Stack Angle: 22% reply rate
  • Three Subject Line Options — Selectable chips tied to each variant angle
  • Brand Compliance Badge — Green "Compliant" or amber "Review Needed" with inline phrase-level flagging
  • Inline Editing — Every email element editable directly in the preview
  • Per-Section Regeneration — Refresh any individual section (Hook / Value Bridge / CTA) without touching the rest
  • One-Click Export — Copy to Clipboard or Open in Salesloft

AI Conversion Intelligence Panel

Expandable panel below each email variant showing:

  • Reply rate — live percentage based on historical sends
  • Meeting rate — percentage of replies that converted to a booked meeting
  • Sample size — number of sends the model has learned from
  • Best fit persona — the role type and situation this angle performs best for

Expanding the panel reveals the top signal combinations the model has learned drive the highest conversion lift — for example:

"Recent LinkedIn post + VP role change < 12 months → +68% lift"
"Q3 earnings beat + CDP gap identified → +54% lift"

Outcome Feedback Row

A lightweight feedback mechanism below every generated email:

  • Meeting booked — logs a positive meeting conversion signal
  • Got reply — logs a positive reply signal
  • No response — logs a negative signal

Every click feeds back into the conversion model in real time. The UI confirms capture immediately — closing the loop between outreach outcome and model improvement with zero friction for the SDR.

Generation Controls (Right Panel)

  • Tone Selector — Formal / Professional / Conversational
  • Length Selector — Short / Medium / Long
  • Active Compliance Rules — Visual display of enforced brand guidelines
  • Regenerate All — Reruns full email generation with current settings
  • Recent Prospects History — Last 5 generated prospects as clickable reload cards

What's Live in V1

Feature Status
Three AI signal stream cards ✅ Live
Live data source status bar (LinkedIn, S&P, BuiltWith, Factiva) ✅ Live
Per-signal source tags, timestamps, confidence bars ✅ Live
Synthesized intelligence summary ✅ Live
Personalized email generator ✅ Live
Three variant tabs with live reply rates ✅ Live
AI Conversion Intelligence panel with signal lift data ✅ Live
Outcome feedback row (meeting / reply / no response) ✅ Live
Three subject line options ✅ Live
Brand compliance checker with inline flagging ✅ Live
Inline editing + per-section regeneration ✅ Live
Tone and length controls ✅ Live
One-click copy to clipboard ✅ Live
Recent prospects history ✅ Live

V2 Roadmap

CRM Sync One-click push of prospect record, generated email, and outcome signal to Salesforce — closing the loop between ProspectIQ and the SDR's system of record without tab switching.

Team Intelligence Library Shared repository of high-performing outreach patterns across the SDR team. Managers identify and scale what works. New reps learn from the team's best emails on day one.

Account-Level Signal Aggregation Expand intelligence gathering from individual contact to full account — multiple stakeholders, org-level signals, buying committee mapping. Enables multi-threaded enterprise outreach from a single account input.

Sequence Generator Extend the single-email engine into a full multi-touch outreach sequence — follow-up emails, LinkedIn messages, and call talk tracks generated from the same intelligence summary, each optimized for the stage of the conversation.


Success Metrics

North Star

Outreach-to-Meeting Conversion Rate — percentage of ProspectIQ-generated outreach sequences that result in a booked meeting.

Supporting KPIs

Metric What it validates Target
Research-to-draft time Is the efficiency problem solved? <5 min vs. 75–90 min baseline
Email acceptance rate Is AI output quality high enough to trust? >65% sent without major edits
Brand compliance pass rate Are guidelines enforced at generation time? >90% from day one
Signal acceptance rate Is the AI finding relevant intelligence? >70% of bullets kept
Outcome feedback capture rate Are SDRs logging results to feed the model? >60% of sends logged within 48h
Conversion lift per signal combination Is the model learning and improving generation? Measurable lift within 60-day cohort
Daily active usage Are SDRs making this their default tool? >80% within 30 days

User Profile

Primary user: Enterprise SDR managing 50–100 named strategic accounts at a large B2B software company.

Profile: Expected to produce deeply personalized, strategically relevant outreach for high-ACV prospects. Currently spending the majority of each working day on research they know is incomplete — pulling from Salesforce, LinkedIn Sales Navigator, investor relations pages, and BuiltWith in separate browser tabs with no synthesis layer.

Core frustration: The research takes so long that cutting corners feels necessary. But cutting corners produces generic outreach. Generic outreach stalls conversion. The loop tightens every week.


Project Structure

artifacts/prospectiq/
├── src/
│   ├── pages/
│   │   └── ProspectIQ.tsx       # Main three-panel layout
│   ├── components/
│   │   ├── DataSourcesBar.tsx   # Live source connection status bar
│   │   ├── IntelligenceCard.tsx # Signal cards with confidence bars
│   │   ├── ConversionPanel.tsx  # AI learning loop + outcome feedback
│   │   └── EmailSection.tsx     # Email body section with regenerate
│   ├── data/
│   │   └── prospect.ts          # Data types, mock intelligence, conversion stats
│   └── index.css                # Tailwind theme + Adobe brand tokens
├── package.json
├── vite.config.ts
└── tsconfig.json

Tech Stack

Built with Replit Agent (vibe coding methodology).

  • Frontend: HTML / CSS / JavaScript — single-page application
  • Live Data Layer: Simulated real-time connections to LinkedIn Sales Navigator, S&P Capital IQ, BuiltWith Intelligence, and Factiva News Feed with realistic latency indicators, timestamps, and confidence scoring
  • Conversion Model: Client-side learning loop tracking signal combinations, reply rates, and meeting rates per email variant — with outcome feedback capture
  • Styling: Enterprise SaaS aesthetic — white background, Adobe red (#E02020) primary CTAs, teal (#0F6B75) intelligence card accents, dark status bar for live data layer
  • No backend required at prototype stage — all state managed client-side

How to Run

  1. Open the live app: (https://prospect-iq-skyplon.replit.app/)
  2. The app loads pre-populated with a completed example (Jennifer Walsh / Workday) — no login required
  3. Observe the live data status bar showing real-time source connections and signal freshness
  4. To generate a new prospect:
    • Enter prospect name, company, role, and outreach goal
    • Click "Generate Intelligence"
    • Review the three signal cards — each showing source, timestamp, and confidence score
    • Select a variant tab — note the live reply rate for each angle
    • Expand the AI Conversion Intelligence panel to see which signal combinations are driving lift
    • Adjust tone and length as needed
    • Click "Copy to Clipboard" to move to your sending tool
  5. After sending, log the outcome using the feedback row — Meeting booked / Got reply / No response — to feed the conversion model

Getting Started

Prerequisites

  • Node.js 20+
  • pnpm 9+

Installation

git clone https://github.com/your-username/prospectiq.git
cd prospectiq
pnpm install

Development

pnpm --filter @workspace/prospectiq run dev

The app runs at http://localhost:<PORT> (port is auto-assigned via the PORT environment variable).

Build

pnpm --filter @workspace/prospectiq run build

Author

Juan Manuel Navarrete Solano
Senior Product Manager — Agentic AI & Generative AI
LinkedIn

About

AI-powered sales intelligence tool for Enterprise SDRs. Aggregates prospect signals across financial reports, LinkedIn activity, and tech stack data — generating personalized, brand-compliant outreach emails in under 5 minutes. Built as a vibe coding MVP prototype using Replit Agent.

Resources

Stars

Watchers

Forks

Releases

No releases published

Packages

 
 
 

Contributors