These 2 job seekers built AI chatbots to talk to recruiters for them – CNBC

Joshua Curry and Vishal Patil have seen a lot of customer service chatbots. The chat windows that pop up on your screen while visiting sites from online retailers to cell phone companies, asking what you need help with, have proliferated in recent years.

Curry and Patil have built similar chatbots. But theirs don’t live on business websites or solve customer service issues. Theirs are on their portfolio websites — and are meant to help them find work.

Curry and Patil’s personalized AI chatbots draw on their application materials and professional experience to interact with recruiters visiting their websites. They hope the chatbots will help land them jobs in today’s tough, “low-hire” job market; U.S. employers added just 116,000 jobs last year, compared to 1.46 million in 2024. Their approach is just one creative solution among many — like sending snail mail or turning to reverse recruiters — that applicants are trying in the hopes of getting noticed.

Curry, a web developer in San Francisco, has gotten interviews for two positions of roughly 140 he’s applied to since November. “It’s insane out there,” he says. “I have to stand out.”

For Patil, an F-1 student visa holder, the job market feels especially daunting, he says. He estimates he’s gotten a screening call or interview for 30 applications of the more than 700 he submitted from mid-August 2025 through March.

Job seekers today face stiff competition: Every minute, approximately 8,200 applications are submitted on LinkedIn, and 38% of U.S. job seekers say they’re applying to more jobs than ever yet hearing back less often, the company told CNBC Make It in April.

Patil initially built his chatbot, VAi, to solve a persistent problem he faced in his search for a summer internship. Juggling coursework for his master’s degree in software engineering at the University of Maryland, College Park, along with his on-campus job in IT, the 26-year-old couldn’t always respond to recruiters as quickly as he wanted. He says he felt even the slightest delay “hampered” his search.

“They have thousands of applications,” Patil says, adding that he felt like “if they don’t get the information immediately, they’ll move on to the next candidate.” He needed to be faster, he thought, and a chatbot could respond instantly.

Curry made his chatbot, ChatJC, after seeing several job posts seeking someone who could build something similar; he wanted to show he could be that person.

Hiring managers in his line of work “want to read the code that you make.” By releasing ChatJC as an open-source project, “you can see my work,” he says, and others can make their own applicant chatbots.

Plus, he says, “I like the idea of something doing work for me while I’m sleeping.”

Both Patil and Curry say it took about two weeks from starting to design their chatbots to deploying them on their websites.

VAi uses information from Patil’s LinkedIn, resume and portfolio website. Suggested prompts like, “Give me a quick summary of Vishal,” or, “What are Vishal’s strongest skills?” can get the conversation started. ChatJC similarly draws on materials like Curry’s resume, cover letters, endorsements and volunteer work.

There were some hiccups at first. On ChatJC, for example, some visitors typed profanity into the question field. Patil and Curry created guardrails to keep their chatbots from offering up sensitive personal information like their home addresses or answering irrelevant questions.

There’s also the matter of hallucinations. Curry has prompts directing ChatJC to only state what it can find in the materials he’s given it. But, as he notes and as LLMs have shown, “chatbots don’t always follow that direction.” Patil likewise has instructed his chatbot to “not go off track” and “not lie” about his experience by, say, claiming he has a skill he doesn’t.

Curry can see the questions people ask but not who asks them. Questions so far include: “What is his tech stack?” (ChatJC answered that well, Curry says.) “What’s Joshua’s main weak point?” (ChatJC didn’t answer.) “Can he code like mad?” He’s had 80 unique visitors since launching in March, according to his analytics.

As for Patil, VAi has received over 3,300 views since launching in January, according to his analytics. In a recent 30-day period, it got 328 unique visits and was asked 492 questions, he says, though he can’t see what those questions were.

Patil says his visibility and reach on LinkedIn has grown from making and posting about VAi, getting more eyeballs on his profile during his job search. “Without [the chatbot], I don’t think that would have happened,” he says. “It has opened up a lot of opportunities for me.”

A handful of recruiters, for jobs Patil had applied to and others, gave him kudos or feedback on VAi, each offering either an interview or technical assessment, Patil adds. Some peers reached out asking how he built VAi, and one person offered to refer Patil for a job.

“I never had these kinds of discussions with any peers or any recruiters” before VAi, he says. “It was all grinding alone.”

“The actual people applying for jobs, we’re lost in the shuffle,” Curry says. As a candidate, he adds, “You’re just sending this stuff out into a void, and nothing shines back.”

ChatJC is his answer of sorts to employers’ use of AI in the hiring process. “I don’t see it as gaming [the system]. I see it as matching it,” Curry says. In “making use of the same tools” on the applicant side, “the goal is to be on equal footing,” he adds. “It’s really just to lift ourselves up and say, ‘Look, we’re not just mice in a cage running mazes.'”

Patil has secured a summer internship; he was offered a continuation of his current work in his university’s IT department, partly, he says, because a colleague saw VAi and was impressed.

For Curry, so far, ChatJC has been a great conversation starter in tech circles and offered useful practice building a chatbot, he says. But his bigger aim with it, as with his other search efforts, is clear: “I hope this damn thing gets me a job.”

Want to get ahead at work? Then you need to learn how to make effective small talk. In CNBC’s new online course, How To Talk To People At Work, expert instructors share practical strategies to help you use everyday conversations to gain visibility, build meaningful relationships and accelerate your career growth. Sign up today! Use coupon code EARLYBIRD for 20% off. Offer valid from April 20, 2026 to May 4, 2026. Terms apply.