Use their tool
Made and published by Ivan Nedelkovski (interviewed by James Fleischmann). Start here; it is their original, not ours.
Our adapted version
Ours Our version, based on the features Lancer lists on its own site (AI job filtering, client-quality filters, proposals from your knowledge base, manual approval). Not Ivan's wording. Fill in the [BRACKETS].
What they did
- Picked a problem he had himself: getting leads on Upwork.
- Built Lancer, an AI agent that finds Upwork jobs for freelancers and agencies and writes the proposals.
- Sold it as a monthly SaaS.
- Reports $10K MRR within 60 days and $20K MRR now, with a two-person team.
- Heads-up: the numbers are self-reported in an Indie Hackers interview, and most of the interview is for IH+ subscribers (free account); the figures above are from the public part. We have no affiliate deal with Lancer.
- Heads-up: Lancer can submit proposals automatically. Check Upwork's terms before automating bids, and use its manual approval mode if you try it.
Summary written by Agent Cookbook. Figures are quoted as the author stated them and are not independently verified. See the original for full context.
Approval rule
Set it up
- Pick a lead source you know well and already struggle with (his was Upwork).
- Try his product, Lancer, to see what the finished version does; turn on manual approval.
- If you'd rather build your own, use the prompt above with your saved searches and profile notes.
- Keep a human on the send step; the agent only drafts.
- Track replies per proposal and tighten the filters weekly.
Common failures
- Auto-submitting proposals risks your account; keep a human send.
- Generic openers; make the first line answer the job post.
- Burning connects on weak clients; raise the hire-rate and spend filters.
Variations
- Same pattern for LinkedIn job posts or RFP sites.
- Package it as a service for other freelancers.
Get the next money recipe. Free.
New recipes every week, plus the top 10 agent setups pack.
Unsubscribe anytime. Privacy