How to Optimize Upwork Profile for Uma AI Recruiter

A working AI agent prompt that audits your profile for Uma's ranking signals, optimizes for machine readability, and monitors your shortlist placement — so clients find you first.

01The Problem

Why Uma Recruiter never shortlists your profile

Upwork's Uma Recruiter auto-shortlists freelancers for Business Plus clients. If your profile is not legible to the AI, you never make the shortlist — and you don't even know why.

Uma reads your profile before any human does

Upwork's Uma Recruiter uses a reasoning layer built on large language models to rebuild job posts as structured briefs, then scores candidates on profile description, demonstrated expertise, past project performance, portfolio depth, availability, and Job Success Score. If your profile is not legible to the AI, you never make the shortlist.

JSS below 90% filters you out of AI discovery

Upwork's own documentation for AI-driven hiring requires a Job Success Score of 90% or higher, relevant skills, good client feedback, and a public profile. Below that line, you are describing yourself out of a channel that didn't exist a year ago. Most freelancers don't know this threshold exists.

Your profile is not optimized for machine readability

Uma picks the past jobs most similar to the current opening and weighs those — five relevant contracts beat fifty unrelated ones. If your contract titles and client feedback do not name the platform or tool explicitly, the AI cannot match you to relevant jobs. Vague profiles get vague matches.

02The How-To

The Uma optimization AI agent prompt, step by step

Copy it once, paste it into Zaira, and it takes over profile optimization: auditing ranking signals, optimizing for machine readability, and monitoring shortlist placement.

Run the Uma profile optimization agent
Access to the @Upwork MCP server. Act as my Upwork profile optimization agent for Uma AI Recruiter. Your goal is to audit my profile for Uma's ranking signals, optimize for machine readability, and monitor my shortlist placement — so clients find me first.

About me

- My niche: [e.g., AI automation, web development, copywriting] - My skills: [e.g., React, Next.js, TypeScript, AI agents, Zapier] - My rate: [e.g., $75/hr or $5000 fixed] - My experience: [e.g., 5 yrs, 20+ projects, JSS 98%] - My portfolio: [e.g., 3 relevant project URLs] - Current JSS: [e.g., 95% / no JSS yet] - Profile completeness: [e.g., 85% / 100%] - Availability: [e.g., available now / not available]

Step 0: Check tools

List the Upwork MCP tools available and how you'll use each (profile read, profile update, job search, shortlist check). Never guess profile data; report only what the tools return.

Step 1: Audit my profile for Uma's ranking signals

Evaluate my profile against Uma's known criteria: 1. Profile description: Does the title and first 2 lines name my exact service and client type? 2. Demonstrated expertise: Do my contract titles and client feedback name the platform or tool explicitly? 3. Past project performance: Are my 5 most recent similar contracts highlighted? 4. Portfolio depth: Do my portfolio items use result-first titles with concrete outcomes? 5. Availability: Is my availability badge accurate and current? 6. Job Success Score: Is my JSS above 90%? If not, flag as priority.
Report specific gaps and recommended changes for each.

Step 2: Optimize for machine readability

Make my profile legible to Uma's AI: 1. Title: '[Primary keyword] | [Secondary keyword] & [Specialty or tool]' 2. Overview first 250 characters: Primary keyword + deliverable + client type 3. Skills: Top 5 = primary keywords, 6-10 = tools and platforms, 11-15 = adjacent skills 4. Portfolio: Result-first titles (e.g., 'Reduced cart abandonment 23% for [Brand]') 5. Contract titles: Ask clients to name the platform or tool in their feedback

Step 3: Optimize past contract discoverability

Uma picks the past jobs most similar to the current opening. To improve matching: 1. Ensure my 5 most recent similar contracts are complete and well-described 2. Ask past clients for specific feedback that names the platform or tool 3. Highlight contracts that match my target niche 4. Remove or minimize unrelated contracts that dilute my profile signal

Step 4: Monitor shortlist placement

Check whether I appear on Uma shortlists for jobs matching my niche: 1. Search for jobs matching my skills 2. Check if I am on the shortlist (if visible) 3. Track shortlist rate over time 4. If shortlist rate is low, revisit Step 1 and Step 2

Step 5: Schedule recurring profile maintenance

Set up a monthly profile review: 1. Update skill tags based on evolving client search patterns 2. Refresh portfolio items with recent results 3. Refine overview based on which profiles actually get shortlisted 4. Small edits signal activity to the algorithm and keep visibility stable
03Why People're Using

What this does once it's running

Three stages, one goal. Zaira runs the optimization pipeline on a schedule, so the only thing left for you is reviewing the changes it made and watching your shortlist rate climb.

Get on Uma's shortlist without bidding

The agent audits your profile against Uma's ranking signals and makes specific changes. Freelancers who optimize for machine readability report appearing on shortlists within weeks — without spending a single Connect.

Make your past contracts work for you

Uma picks the past jobs most similar to the current opening. The agent ensures your 5 most recent similar contracts are complete, well-described, and highlighted. Five relevant contracts beat fifty unrelated ones.

Unlock a discovery channel most freelancers don't know exists

Uma Recruiter launched in October 2025 for Business Plus clients. Clients who use it hire 30% more often. The agent optimizes your profile for this channel — so you get shortlisted for jobs you never even see.

04FAQ

Frequently asked questions

The practical questions people ask before optimizing for Upwork's AI recruiter.

Uma Recruiter is Upwork's AI agent that builds a client's shortlist. When a client requests shortlisting on a public job post, Uma reads the post, searches the talent pool against stated and inferred requirements, scores candidates on profile, past work, portfolio, availability, and JSS, invites the strongest to apply, and ranks incoming proposals. The shortlist is typically ready within 6 hours.

Upwork's documentation specifies: Job Success Score of 90% or higher, relevant skills, good client feedback, and a public profile. Beyond the hard gate, the same signals that win regular search matter: your skills list must match what a client would type, and your work history needs to read as real proof, not filler.

Regular profile optimization targets human readers and Upwork's search algorithm. Uma optimization targets an AI that reads your profile before any human does. The agent ensures your profile is machine-readable: clear skill tags, result-first portfolio titles, specific client feedback that names platforms, and contracts that match your target niche.

Yes, if the shortlist is visible to you. The agent checks your shortlist placement for jobs matching your niche and tracks your shortlist rate over time. If your shortlist rate is low, the agent revisits your profile optimization and identifies gaps.

With a complete, machine-readable profile and JSS above 90%, most freelancers see shortlist placement within 2-4 weeks. The key is specificity — 'Shopify Developer for DTC Apparel Brands' matches exact client searches, while 'Web Developer' competes with 200,000 profiles.