High-ranking executive lead gen for African corporations
# exa-lead-agent-hnw
AI-powered lead generation pipeline for high-net-worth executive prospecting across Sub-Saharan Africa. Targets C-suite, VP-level, and Director-level officers at major SSA corporations for luxury real estate outreach.
Built on Grok 4.20 (xAI) for executive discovery and Exa semantic search for verification and enrichment. Delivers ICP-scored, deduplicated lead lists with LinkedIn URLs, role tiers, confidence scores, and contact routes — ready for direct outreach.
---
## SSA Executive Leads (xAI)
**Current (v4):** Grok-led pipeline + capped Exa — debug/UI artifacts under `outputs/pipeline/`, plus a legacy-compatible `jsons/*.enriched.json` per run and a locked refresh of aggregate files under `fullJSONs/`.
```bash
pip install -r requirements.txt
export XAI_API_KEY=...
export EXA_API_KEY=...
python -m pipeline run
example-corp.com
```
To skip aggregate writes for experimentation:
```bash
python -m pipeline run
example-corp.com --no-aggregate-sync
```
**Legacy scripts (archived under `legacy/`):** root shims keep old commands working; implementation files live in `legacy/`.
1. **Research** — `hotel_decision_maker_research.py` (shim → `legacy/hotel_decision_maker_research.py`) discovers executives at a target corporation, writes JSON under **`jsons/`** and appends CSV under **`csv/`** (needs `XAI_API_KEY`).
2. **Contact enrichment** — `hotel_contact_enrichment.py` (shim → `legacy/hotel_contact_enrichment.py`) re-reads that JSON, runs `grok-4.20-reasoning` with web + X search per candidate, merges email/phone/X/LinkedIn back in. Skips rows that already score high on direct channels.
```bash
pip install -r requirements.txt
export XAI_API_KEY=...
python hotel_decision_maker_research.py --url
example-corp.com
python hotel_contact_enrichment.py --in-json jsons/corp_leads__....json --out-json jsons/corp_leads__....enriched.json
python hotel_contact_enrichment.py --in-json in.json --out-json out.json --dry-run
```
### …