Donna AI: Hiring via AI Agents—A Bold Bet by an Indian Student Team

2026-03-13 | Product Hunt | Official Website

30-Second Quick Judgment

What is it?: Both the recruiter and the candidate get an AI Agent. These two agents communicate continuously in the background to align expectations. Only highly matched candidates ever appear before you. Essentially, it turns 'networking referrals' into a 24/7 automated AI process.

Is it worth watching?: The concept is sexy—'agent-to-agent' is one of the hottest narratives for 2026. However, the product is in its infancy (still in the 'Get early access' stage), founded by two Indian students applying for YC. It's great for tracking industry trends, but not yet ready for heavy-duty use.

Three Key Questions

Is it relevant to me?

• Target Audience: Early-stage startup founders, HR leaders (especially those who hate screening resumes), and candidates who want to avoid mass-applying.
• Am I the target?: If you're a small team founder who relies on friend referrals but feels the 'pool isn't big enough,' you are. If you're a candidate tired of your resume falling into a black hole, you are too.
• When would I use it?:
  • Hiring the first 10 people for a startup where culture fit is key but your network is limited → Use this.
  • Passive job seekers who want opportunities to find them → Use this.
  • Standardized, high-volume hiring for large corps → Not suitable; traditional ATS is more mature.

Is it useful?

ROI Judgment: It's too early to jump in fully. The product is in early access, and a small user base means a small candidate pool. Track the concept, but don't expect to hire your next star today.

Is it delightful?

The Highlights:

• Intuitive Agent Interface: Recruiters get an agent named 'Jessica,' and candidates get a 'Career Agent.' Both interact via Telegram, making it feel like texting a recruiter friend.
• Salary Pre-alignment: The most annoying thing is finishing three interview rounds only to realize the salary doesn't match. Donna's agents solve this before the first interview.

The 'Wow' Moment:

"Our eventual goal with Donna is to create a platform where your digital twin can network for you." — @furst_fly (Founder)

This quote highlights the true vision: it's not just about hiring; it's about letting your digital twin handle your social networking. That's a massive space for the imagination.

Real Feedback:

"Sourcing high-intent candidates is the hardest part. Even with many applicants, the right-fit people rarely appear in the pipeline." — Mukund Shah, Founder, Emergent AI

"Even agencies don't fully understand our hiring taste. Their 7/10 candidate is often a 5/10 for us." — A D2C Startup Founder

(Note: These quotes are from the website's validation section, collected during founder research, not actual Donna user reviews.)

For Independent Developers

Tech Stack

• Frontend: Web + Telegram Bot (Main interaction entry)
• Backend: Undisclosed, features an API interface ("Your next great hire is one API call away")
• AI/Models: Undisclosed, uses an Agent architecture for continuous bilateral matching
• Infrastructure: Includes a Live Hiring Map for real-time dynamics

Core Implementation

The core is a two-layer Agent system: The Recruiter Agent learns the founder's 'hiring philosophy' (a clever term—focusing on subjective preference rather than just hard metrics), while the Career Agent tracks what the candidate actually wants. The two agents talk 24/7 in the background, scoring and aligning on salary and role expectations. Only candidates with a high enough alignment score are pushed to the interview stage.

The challenge isn't the LLM calls; it's:

1. How to accurately learn 'subjective preferences'—every founder's definition of 'culture fit' is different.
2. The cold start problem—agent matching is meaningless if the candidate pool is too small.
3. The chicken-and-egg problem of a two-sided market.

Open Source Status

• Closed source, no related code on GitHub.
• Similar Open Source Projects: No direct open-source agent-to-agent hiring competitors yet.
• Build Difficulty: A prototype of the core architecture could be built in 2-3 months, but the supply side (candidate pool) is the real moat.

Business Model

• Monetization: Success Fee (Commission on successful hire).
• Pricing: Explorer (Free) / Success Fee (15-20% of annual salary) / Enterprise (Custom).
• User Base: Undisclosed; based on PH votes and Twitter buzz, it's very early.

Giant Risk

High Risk. LinkedIn has already launched Hiring Assistant—the first official AI recruiter agent. With over 1 billion users, LinkedIn has a natural advantage in agent matching. Mercor has already reached a $10B valuation with a $350M Series C. This sector is not lacking in cash or giants.

Donna's moat must be 'understanding startup hiring better than the giants' or 'perfecting the agent-to-agent bilateral experience.' Neither is fully established yet.

For Product Managers

Pain Point Analysis

• Problem Solved: The 'noise' in hiring—massive volumes of mismatched applications make screening costs sky-high, while the best candidates are hidden in limited personal networks.
• Severity: High-frequency, critical need. 'Screening takes hours per hire' is a daily reality for startup HRs and founders.

User Personas

• Persona 1: Early-stage startup founders (10-50 people) who are personally involved in hiring and have strong 'culture fit' preferences that are hard to scale.
• Persona 2: High-quality passive job seekers who don't want to mass-apply but want to be matched with the right roles.

Feature Breakdown

Competitor Comparison

Key Takeaways

1. The 'Hiring Philosophy' concept is great—structuring subjective preferences is closer to real decision-making than traditional JD filtering.
2. 'No applications → no noise'—eliminating noise at the source rather than filtering through it is a clear, smart strategy.
3. Success Fee Pricing—aligns interests and lowers the barrier for users to try it out.

For Tech Bloggers

Founder Story

• Founders: Dhruv Agarwal (@furst_fly) and Dawar (@Dawar1213).
• Background: Indian student entrepreneurs.
• A Great Anecdote: At the AI India Impact Summit, they couldn't afford a ₹30,000 (approx. $360) booth. On day one, they scouted empty booths as 'student journalists.' On day two, they printed a ₹100 poster and occupied a canceled booth for three days. In their words: "This is our answer to the YC hack question."
• Currently applying for Y Combinator.

Controversies / Discussion Angles

• Is 'AI Agents socializing for you' too sci-fi? The vision of a 'digital twin networking for you' touches a sensitive nerve: Can human relationships be delegated?
• Transparency in Bilateral Agent Negotiations: When two AIs are negotiating your salary or assessing 'culture fit,' do you know what they're saying? Is black-box decision-making a bug or a feature?
• Students vs. $10B Mercor: The David vs. Goliath narrative always gets clicks.

Hype Data

• PH Ranking: Low vote count (16 votes / badge shows 121), not a 'hot' launch.
• Twitter Buzz: Very few related tweets, mostly from the founders themselves.
• Search Trends: The name 'Donna AI' is used by at least 4 different products (hiring, legal, music, sales), making SEO very messy.

Content Suggestions

• Angle: "When AI Agents start socializing for us, hiring is just the first step"—use Donna as a hook to discuss the agent-to-agent trend.
• Trend Opportunity: YC application season + the agent narrative. Write about the '₹100 poster' startup story.

For Early Adopters

Pricing Analysis

Getting Started

• Setup Time: Approx. 10-15 minutes.
• Learning Curve: Low—interact via Telegram; no new UI to learn.
• Steps:
  1. Visit trydonna.net and click 'Get early access' to register.
  2. Choose 'I'm Recruiting' or 'I'm Looking for a Job.'
  3. Chat with your Agent on Telegram to tell it your preferences.
  4. Wait for the Agent to match and push results.

Pitfalls & Critiques

1. Extremely Early Stage: The About, Blog, Docs, and Careers pages are all empty links (#), indicating the infrastructure isn't ready.
2. Small Candidate Pool: As a new two-sided market, there aren't many people on either side yet, so matching quality is unproven.
3. Generic Brand Name: 'Donna AI' is shared by at least 4 companies. Finding specific product info is a chore.
4. No Independent Reviews: No feedback found on Reddit or Twitter from non-founders.

Security & Privacy

• Data Storage: Specifics undisclosed.
• Privacy Policy: The Privacy page is an empty link.
• Security Audit: None.

Alternatives

For Investors

Market Analysis

• Core AI Recruitment Market: ~$656M (2024) → $1.2-1.4B (2033-2035), CAGR 6.8-7.4%.
• Broader AI in HR Market: $6.25B (2026 est.), CAGR 24.8%.
• Agentic AI Sector: $5.2B (2024) → $196.6B (2034), CAGR 43.8%.
• Drivers: 87% of companies are using AI hiring tools in 2026; 93% of recruiters plan to increase AI usage.

Competitive Landscape

Timing Analysis

• Why Now?: 2026 is the turning point for AI agents moving from concept to reality—agent-to-agent protocols (like MCP) are maturing, making 'two agents talking' technically feasible.
• Tech Maturity: LLMs are now capable enough for preference learning and matching; Telegram bot development costs are low.
• Market Readiness: Acceptance of 'AI agents doing things for me' is rising, though 'AI agents socializing for me' still needs user education.

Team Background

• Founders: Dhruv Agarwal + Dawar, Indian student entrepreneurs.
• Core Team: Size undisclosed, estimated 2-3 people.
• Track Record: First-time founders, no known previous track record.

Funding Status

• Raised: Undisclosed (likely bootstrapped).
• Applying for Y Combinator.
• Valuation: Unknown.

Conclusion

Bottom line: A sexy concept but extremely early—'agent-to-agent hiring' is a great story, but Donna AI has a long way to go to tell it well.

Resource Links

2026-03-13 | Trend-Tracker v7.3 | Data Sources: WebSearch, Twitter/X, Playwright Web Scraping