← All episodes

I Built a $10K/Month AI Agent

| 15 products mentioned
Watch on YouTube ai agents platform automation saas growth affiliate marketing upwork automation ai product development startup strategy

Ivan Nedlekovski built Lancer, an AI agent that automates job discovery and bidding on Upwork, reaching $10,000 monthly revenue in just 3-4 months without paid advertising. The episode explores his unconventional growth strategy using "connectors"—identifying influential intermediaries (Upwork coaches) with trusted networks rather than building an audience from scratch, demonstrating a replicable playbook for platform automation AI products.

Key takeaways
  • AI agents automating tedious workflows on existing platforms like Upwork represent a massive opportunity, as they solve real pain points that previously required 10+ hours per week of manual work.
  • The "connector" strategy—identifying and recruiting trusted influencers or coaches within your target market as affiliates—bypasses the need to build an audience and provides instant access to qualified customers through high-trust relationships.
  • Most growth (two major affiliates) came from personal referrals and direct outreach, with commission structures of 20-30% lifetime recurring revenue proving sufficient incentive for high-value partners to actively promote the product.
  • The tech stack leverages TypeScript, Next.js, Node.js, and modern AI APIs to build quickly, with Claude Opus and Cursor enabling rapid development at minimal engineering overhead.
  • Defining your Ideal Customer Profile (ICP) first—rather than building features speculatively—is critical; Lancer's ICP is agencies doing high-volume proposals, not individual freelancers, which shaped product positioning and affiliate targeting.
  • Any platform with a large user base experiencing repetitive, tedious workflows is a candidate for AI automation; examples include Fiverr, LinkedIn, Poshmark, and Pinterest beyond Upwork.

Recommendations (12)

TypeScript
TypeScript uses

"TypeScript on both the front end and the back end using Next.js and Node.js"

Ivan Nedlekovski · ▶ 5:17

Upwork
Upwork uses

"Our main client recruiting channel is Upwork, which turned out to be a very underrated source of quality leads."

Ivan Nedlekovski · ▶ 3:03

Next.js
Next.js uses

"TypeScript on both the front end and the back end using Next.js and Node.js"

Ivan Nedlekovski · ▶ 5:21

Node.js
Node.js uses

"TypeScript on both the front end and the back end using Next.js and Node.js"

Ivan Nedlekovski · ▶ 5:21

Google Cloud Platform

"GCP and Firestore for the hosting"

Ivan Nedlekovski · ▶ 5:23

Firestore
Firestore uses

"GCP and Firestore for the hosting"

Ivan Nedlekovski · ▶ 5:26

Cursor
Cursor uses

"At this point we're all using Cursor with Opus 4.5 and we barely touch any actual programming language"

Ivan Nedlekovski · ▶ 13:08

Claude Opus

"At this point we're all using Cursor with Opus 4.5 and we barely touch any actual programming language"

Ivan Nedlekovski · ▶ 13:08

OpenRouter
OpenRouter uses

"We use OpenRouter for the APIs for the LLMs"

Ivan Nedlekovski · ▶ 13:16

Render
Render uses

"For hosting we use a combination of Render and GCP"

Ivan Nedlekovski · ▶ 13:18

Elasticsearch

"We use Elasticsearch for basically querying the jobs and all the data we store"

Ivan Nedlekovski · ▶ 13:27

Rewardful
Rewardful uses

"We use this software tool called Rewardful. It's affiliate marketing software. It's pretty great and that makes our life very easy."

Ivan Nedlekovski · ▶ 12:07

Mentioned (3)

ChatGPT
ChatGPT "I would go to ChatGPT and look up platforms similar to Upwork" ▶ 12:31
Loom
Loom "Do even like a long Loom video or reference again like some of their work" ▶ 11:02
OpenAI API
OpenAI API "We actually debated with my co-founder to go all in on AI products even back when OpenAI released..." ▶ 14:26