Ajay Dabas, a business owner in Delhi, wanted to know whether AI could take the grind out of cold outreach. Finding the right people on LinkedIn, collecting their contact details and writing a message that doesn't read like spam can take hours for every batch of prospects.
In two weeks, I built a working prototype in Python. You pick a country and an industry, describe your business and the problem you solve, and the tool returns a ready-to-use spreadsheet of prospects, each with a personalised outreach email written by GPT.
Client: Ajay Dabas, Delhi, India
My role: Python & AI developer
Timeline: 2 weeks
Stack: Python, OpenAI GPT-4o mini, Proxycurl API, asyncio, aiohttp, Jinja2
Results
Hours of research cut to minutes: a list of matching prospects, with contact details, in a single run.
1 personalised email per prospect, addressed by name and written around the client's business and the prospect's likely challenge.
Profiles fetched in parallel, so larger lists take barely longer than small ones.
1 spreadsheet, ready to use: profile links, names, emails and messages that drop straight into a CRM or email tool.
A working proof of concept in 2 weeks, with a clear path to a product.
The challenge
For a small business, LinkedIn is one of the best places to find decision-makers, but turning it into leads is mostly manual: searching profile by profile, hunting for a way to reach each person, and writing messages that don't get ignored.
What I built
A guided, keyboard-driven menu. Choose the target country and industry with the arrow keys, then describe your business and your prospects' likely challenge. That's all the input it needs.
Prospects and contact details via APIs, not scraping. The tool uses Proxycurl's people-search and profile APIs for structured data, with profile lookups running concurrently through asyncio.
Personalised emails written by GPT. GPT-4o mini drafts an outreach email from your business description and the prospect's likely challenge. To keep it reliable, the model gets strict rules (no subject line, no HTML, a professional tone and just 3 placeholders) and must answer inside a marked block that the tool extracts and fills in with Jinja2. A low temperature keeps the wording consistent from run to run.
A timestamped CSV for every run, with each prospect's LinkedIn profile, name, email addresses and personalised message.
Read the full case study, with the complete architecture and more screens, on my website: akshatarora.in/work/ai-linkedin-lead-generator
0 likes · 0 views · Nov 11, 2026
A 2-week Python + GPT prototype that cuts LinkedIn prospecting from hours to minutes: prospects found by country and industry, contact details fetched in parallel, and a personalised email for each one, in a single CSV.
