AboutExpertiseProjectsJourneyBlogContact
Discuss
AboutExpertiseProjectsJourneyBlogContactDiscuss

Yao David Logan

Software Engineer fullstack specialized in SaaS, business automation and scalable web/mobile platforms.

NavigationExpertiseProjectsJourneyBlogContact
LinksGitHubLinkedInEmail
© 2026 Yao David Logan. All rights reserved.
← Back to projectsFullstack

Prospect-Pro AI

AI-enhanced B2B prospecting: multi-channel sourcing, lead scoring and message personalization.

Architecte logiciel & Fullstack Engineer202660% higher conversion rate, ~15 hours saved per week
Context2026

AI-enhanced B2B prospecting: multi-channel sourcing, lead scoring and message personalization.

B2B prospecting SaaS that automates sourcing, scoring and outreach. The product ingests company contacts from multiple sources (Pages Jaunes, Pappers, GoAfrica, Google Maps), enriches the data through OpenAI and generates personalized messages that drop directly into client CRMs. Built for agencies and B2B sales teams who want to move from manual sourcing to an automated chain — without losing targeting quality or the legal compliance of the scraping pipeline.

Scope
Architecte logiciel & Fullstack Engineer · Fullstack
Stack
Next.js · Node.js · TypeScript
Signal
60% higher conversion rate, ~15 hours saved per week
ChallengeBefore intervention

Blocked identified.

Manual B2B prospecting in the West African context suffers from scattered contact bases, slow enrichment and generic messages that end up in the trash. Sales teams spent more time building Excel files than talking to qualified leads.

Risk addressedArchitecte logiciel & Fullstack Engineer · Fullstack
ApproachAfter delivery

Answer built.

Modular ingestion pipeline: each source (Pages Jaunes, Pappers, GoAfrica, Google Maps) is wrapped in a typed connector with rate-limiting and robots.txt compliance. Contacts are normalized into a single schema stored in PostgreSQL. Enrichment layer: OpenAI scores leads against client criteria (size, sector, buying signals) and personalizes outreach messages. Native CRM integration via REST API and idempotent webhooks. UI side: a Next.js dashboard with dynamic filters, a Kanban view of the prospecting pipeline and multi-format export (CSV, Excel, JSON).

Foundation deliveredNext.js · Node.js · TypeScript
Results

Measurable impact

Metrics give a quick read of the case study effects.

AI Precision++84% vs before
92%
Before50%
After92%
Qualified leads / month++325% vs before
850
Before200
After850
ROI commercial++192% vs before
3.5x
Before1.2x
After3.5x
EvolutionPost-launch progression
8504250
S1120
S2340
S3580
S4720
S5850
Architecture

System delivered

A simple read of the functional blocks and their interactions.

clientDashboard Next.js
gatewayAPI Node.js
servicePipeline scraping
serviceEnrichment service
serviceService messages
databasePostgreSQL
databaseRedis (queues)
aiOpenAI
externalPublic sources
externalCRM clients
Dashboard Next.jsAPI Node.jsREST
API Node.jsPipeline scrapingJobs
Pipeline scrapingPublic sourcesHTTP rate-limited
Pipeline scrapingPostgreSQLNormalisation
API Node.jsEnrichment servicePipeline
Enrichment serviceOpenAIScoring
Enrichment servicePostgreSQLUpdate
API Node.jsService messagesCompose
Service messagesCRM clientsWebhooks
API Node.jsRedis (queues)Tails
Accuracy scoring92%
Vitesse pipeline88%
Source coverage85%
Customization90%
Adoption client82%
Main signal60% higher conversion rate, ~15 hours saved per week

A 60% increase in conversion rate was observed, and sales teams gained approximately 15 hours per week.

Deliverables
  • Automated sourcing from 4 public sources (Yellow Pages, Pappers, GoAfrica, Google Maps)
  • AI lead scoring based on the client's business criteria
  • Automatic personalization of contact messages (templates + OpenAI)
  • Direct integration into customer CRMs via REST API + webhooks
  • Multi-format export and complete pipeline history
Stack
Next.jsNode.jsTypeScriptPythonPostgreSQLOpenAIPuppeteerRedis
Public demoDiscuss a similar need
Collaboration

Let's turn a fuzzy need into a shippable product.

Product diagnosis, SaaS architecture, backend, interface and automations that make a platform usable.

Format
Full-time, freelance, long mission
Focus
SaaS, API, back-office, automation
Discuss a projectDownload CV