Kevin Vo

I’ve been spending a lot of time mapping agentic workflows in recruiting, here are a few ideas I've built that recruiters and agencies can deploy quickly using Gumloop: Resume Scoring Agents ingest a JD + resume and return a fit score, gaps, and red flags. • Faster screening • More consistent signal • Less resume fatigue Post-Interview Debrief Agents Automate and summarize interviewer calls/notes, resumes, LinkedIn, GitHub, and portfolios into a clean candidate briefs. • Faster processes/candidate hand-offs • Structured/uniform ATS notes and profile submissions Recruiter Copilot (Slack / Browser) Ask: “Should I move this candidate forward?” or “What risks should I probe?” • Helps junior recruiters think like seniors • Reduces decision paralysis Hiring Manager Signals Detects shifting requirements and inconsistent feedback. • Fewer moving goalposts • Less wasted time Agency Ops & Revenue Forecasting Track deal health, probability, and at-risk searches automatically. • More predictable revenue • Better founder visibility Why these work: *Most of these workflows can be built and operational in days. Compounding value fast because they remove repetitive judgment and coordination work. The future of recruiting isn’t fully autonomous AI. It’s AI-powered recruiters who move faster, think clearer, and close better. If you’re building or experimenting with agentic recruiting workflows, I’d love to compare notes! #gumloop #recruiting #ai

4

1 Comment

$6M for one engineer. $5B on GPUs. Welcome to the arms race. AI breakthroughs are bottlenecking on talent and Big Tech is treating top hires like free agents. ➠ Meta’s newest AI hires: $6.2M comp ➠ OpenAI’s compute bill: $1.5B+ per year ➠ NVIDIA’s top clients are spending more on H100s than full ML teams When $30M training runs are routine, $5M salaries are just the premium on performance insurance. But here’s the flip side: → A barbell market: elite talent soars, everyone else gets frozen out → One hire can cost more than 10 teammates, and morale can pay a price The next edge isn’t salary; it’s ownership, autonomy, and IP upside. 📌 What’s your take: smart allocation or dangerous precedent?

45

38 Comments

been thinking...MCP (Model Context Protocol) changed coding agents by giving them crisp, machine-digestible context. Recruiting needs the same thing. Introduce the JCP: Job Context Protocol. It’s the structured layer that turns a job description into context that sells. Clean, modular, and reusable across every candidate touchpoint. So what’s in a JCP? 1-Role Intent: Why this role exists and what “great” looks like in 90 days. 2-Non-Negotiables: Skills, clearances, certifications..no guesswork. 3-Performance Levers: KPIs, core deliverables, and success metrics. 4-Constraints: Location, shift, travel, clearance, on-call. 5-Team DNA: Tech stack, culture signals, decision velocity. 6-Career Vector: Promotion path, exposure, impact narrative. 7-Why Now: Business trigger, urgency, and compelling timing. 8-Comp Levers: Bands, equity/bonus logic, levers that can move. Now make it interactive. How? Use Inbound Voice AI = the MCP layer for candidates. 1-It loads the JCP and does what great recruiters do at scale: 2-Personalized pitch: Tailors the value prop to a candidate’s background in real time. 3-Objection handling: Answers comp, tech stack, visa, clearance, and manager questions 24/7. 4-Depth probes: Dynamic screening that adapts to answers (not a rigid script). 5-Memory & notes: Summaries pushed back to the recruiter with red/green signals. 6-Multilingual reach: Speak to talent where they are—literally and figuratively. 7-Nudge sequences: Follow-up SMS/email with “day-in-the-life,” team videos, and role FAQ. Result: You stop “reading a JD” and start selling a mission. Recruiters become closers. Candidates get instant clarity. Time-to-submissions drops; interview quality rises. At #Giftem, we treat the JD as a JCP and let our Inbound Voice AI run it like an interpretation layer: same source of truth, consistent narrative, compound learning. If you’re still shipping PDF job descriptions, you’re sending agents into production without an MCP. #Giftem #Inboundvoiceai #MCP #JCP #hiring #recruiting #talentacquisition #HRIS #healthcare #Govcon #Govtech

11

Anthropic just dropped Claude Sonnet 4.6, and it’s officially time to rethink your AI strategy. For a long time, the rule of thumb was: "Sonnet for speed, Opus for the heavy lifting." That line just got incredibly blurry. Sonnet 4.6 isn't just an incremental update; it’s a full-scale upgrade that rivals the flagship Opus 4.6 while staying at the Sonnet price point. We are moving away from AI as a "copy-paste" assistant and into the era of Autonomous Workflows. Whether you are building complex agents or just trying to ship code faster, Sonnet 4.6 is the new gold standard for what a "mid-tier" model can do

3

I've talked to a lot of internal TA people over the years. When Internal TA teams are working with an agency, they just want to partner with external recruiters who are easy to work with. They want the necessary support and not an extra headache. Some agencies will try to bypass the process, throw resumes at job reqs they weren't even assigned to, and make things more difficult. Causing more harm than good. It's is a good way to blow it. Be cool. Be easy to work with. The opportunities will be there if you deliver results and not try to force your way into every opening right from the jump. Getting a contract doesn't mean it's a free for all. Not every position needs attention. You can inquire about other openings but submitting a bunch of candidates without a prior discussion is a recipe for disaster.

8

You don’t need more recruiting agencies, you probably don't need a huge internal TA team yet either.. You need an Embedded UA Unit that gets it done. Here’s the reality for a lot of companies: 📌 5–10 open roles that just… linger. Sometimes just a few tough ones. 📌 An internal team juggling too many reqs, triaging every day. Or no internal team yet period. 📌 The temptation to bring in contract recruiters on 6-month engagements, or stack 5 agencies spraying the same Indeed resumes at you. Stop. You don’t need to spin up laptops, onboard recruiters, and hope they understand your business.(then try to get the laptops back after) You don’t need to burn months managing people you hope will figure it out. And you don't need to burn a bunch of real people looking for jobs because the people front running your brand in the employment world aren't getting it. You need an Embedded UA Unit: ✅ Embedded full-desk recruiters from Underground Admins ✅ Already trained. Already working together ✅ Plug in. Own the search. Fill the seats ✅ No handholding. No overhead. No guessing Underground Admins drops in as an extension of your team with zero ramp-up. We own the outbound. The follow-up. The close. We keep your revenue and ops leaders moving while you focus on what’s next. Everything goes into your ATS(we literally have worked on all of them, unless it's a self build) And if you’re not ready for an Embedded UA Unit? We still crush high-stakes one-off searches on a contingent basis. But for companies tired of recruiter roulette, the Strike Team model changes the game. Check our track record. Don’t take our word for it. Click on to my profile. Scroll down. Read the recommendations: ✅ From hiring managers who trusted us with critical seats and got results. ✅ From candidates we placed, people still checking in years later, sending referrals, and asking for career advice. This isn’t luck. It’s reps. It’s follow-through. It’s Underground Admins doing the work while others are still talking about all the unicorns and smurfs and other imaginary creatures they have on their roster in the sweet ChatGPT outreaches they send which usually includes some variation about how they find top tier talent not on the job boards with their magical tools(hint, it's LinkedIn recruiter) DM me. Skip the pain. Get results faster.

17

1 Comment

H-1B Sponsorship Just Got a Lot More Expensive: What It Means for Buy-Side Tech Hiring A new rule adds a $100,000 fee to new H-1B petitions filed after Sept 21, 2025. Current visa holders and routine renewals will not be affected, but any fresh sponsorship will be a lot more expensive. How it hits investment-management tech teams: 1. Tighter pipelines: Mid-sized asset managers and trading firms that used to bring in H-1B engineers will think twice before sponsoring. Talent pipelines will thin at the margin. The new fee forces a rethink of the tradeoff between hiring internationally vs. hiring locally or contracting. 2. Slower, pricier hires: Expect longer searches due to the tighter market of qualified US Citizens or GC holders. This will squeeze firms to rethink how talent pipelines are built and make a decision between cost and time. 3. Vendor squeeze: Big consulting firms might eat the cost, but smaller quant/fintech vendors will likely pass it along. Tips for candidates -Not work authorized yet? Target companies with a track record of sponsoring and be ready to share your immigration timeline. -Already on H-1B? Your renewals inside the U.S. should be fine, but double check before international travel. Ideas for hiring teams: -Know your hiring plan early as you will now need to begin pipelining candidates earlier for planned hires.   -Get on campus early: stronger internship programs and university relationships will help you tap U.S. work authorized grads before they hit the open market. -Recheck vendor contracts and budgets so you know who’s covering the extra cost. -Look at remote or near-shore options if overseas talent still makes sense. Rules and guidance are still shifting, so both candidates and employers should stay in touch with immigration counsel and start planning now.

25

The biggest complaint in the job market is that applications go into a black hole. It's been a growing problem since one-click applications became the norm. Recently, AI has been blamed as the cause, or at least as a major contributing factor, but I believe it's the way out of this mess. We recently started using a solution from Tenzo AI to solve it. We call our new AI screening assistant Ruby, and here's what she does: 🔹Responds within minutes of receiving an application 🔹Evaluates qualifications against the specific role requirements 🔹Offers a chance to go beyond the resume in a screening interview 🔹Schedules qualified candidates to speak with a recruiter the same day 🔹Tells applicants who don't meet minimum requirements exactly where they fell short, and gives them a chance to clarify That last point is my favorite feature. No more waiting and wondering. Every candidate application is screened consistently without bias. Ruby doesn't make hiring decisions; every candidate who moves forward speaks with a human recruiter who knows the client and the role. She handles the volume so our team can spend their time where it actually matters. When's the last time you applied for a job and heard back the same day, whether the answer was yes or no? Our goal is to normalize exactly this. You can learn more about Ruby & why we chose Tenzo as our partner here: https://lnkd.in/eb2gxbg9

26

7 Comments