SelectPrism helps hiring teams meet their strongest candidates without getting buried in the process of finding them. Upload a job description, and SelectPrism takes it from there. It interviews applicants, evaluates their fit, and hands your team a shortlist of people worth meeting. Your recruiters get their time back to focus on what actually moves the business: closing the right people.
Maker
📌Hey Product Hunt 👋
I'm Dr. Rishi Thussu from SelectPrism. I spent years at Monster watching millions of applications pile up and go nowhere. Then at Findem, I was building AI tools for talent intelligence — and seeing what was actually possible when you applied real machine learning to this problem.
Both taught me the same thing: the problem was never finding applicants. It was what happened after.
SelectPrism is an AI interviewer that runs your entire first round, from screening to scoring, so your team shows up only for the conversations that matter. It plugs directly into your existing ATS, so every candidate, report, and hiring decision stays in one connected flow.
We built this for Talent Acquisition (TA) leaders who are great at their jobs but spend most of their week on work that no longer requires a human.
If you're losing good candidates to slow pipelines, this is for you.
The problem we kept running into was simple:
The best candidates don't wait two weeks for a callback. By the time most TA teams finish screening and scheduling, the people worth hiring have accepted other offers.
That's why SelectPrism is built around one complete flow: screen → interview → shortlist.
Here's what that looks like in practice:
📋 Drop in a job description. SelectPrism parses it, maps it to a skill graph trained on 60M+ profiles, and ranks your applicants by actual fit, not keyword matches.
📩 Interview invites go out within the hour. No scheduling, no back-and-forth. Candidates get a link and show up when they're ready.
🎥 The AI conducts the interview. Live video, adaptive questions that adjust in real time, voice-based pre-screening, and coding assessments where the role needs them.
🔒Holistic anti-cheating checks. 25+ integrity checks across browser, device, and video — including gaze detection, lip sync analysis, AI voice agent detection, candidate ID verification, and real-time fraud flagging.
📊 Your recruiter gets a decision-ready report. Fit score, strengths, gaps, full transcript, and a hiring recommendation. Everything needed to decide in five minutes.
Why TA teams switch to SelectPrism:
✅ One platform replaces your screening tool, interview scheduler, assessment software, and proctoring suite
✅ Recruiters stop managing calendars and start closing offers
✅ Every candidate gets a fair, consistent evaluation regardless of when they apply
✅ Hiring managers only meet the people who actually cleared the bar.
We're excited to share this with the PH community and would love your honest feedback. We'll be here all day answering questions.
🎁 Product Hunt Launch offer: 50 free AI interviews!
👉 Try SelectPrism: https://selectprism.ai/agentic-ai-interview
Thanks for checking us out, and huge thanks to our hunter @chrismessina for hunting us 🙏
Report
@chrismessina @thecleric The anti-cheat list is impressive. The harder question for me is on the other side: when a strong candidate gets filtered out by the system (the 6% inaccuracy), is there a feedback loop where a human recruiter can flag "this rejection was wrong" and have the model adjust?
Most hiring tools I've used treat the false negative as invisible because the candidate never knows they were rejected. Curious how you handle the audit side.
Report
Loved the design and UX from the website, but had few more queries.
Do i need a separate ATS, or does this already have an ATS inside it? (to track the status of candidates)
Will these free 50 interviews have all the features enabled, or will their features be locked behind other tiered pricing?
Will there be bulk export options for all the interviews that we take from this platform?
Is it possible right now or in future, if i can just upload the resumes in my chat (be it slack or teams) and just ask the bot to match it against specific job that i have already posted on the platform?
Will all my jobs be public like naukri.com or will the job applications/interview would be invite only? (like i don't want it to be public as there can be too many applications)
How will the scoring work? Like, is there any way i can change the weightage of scoring, different HR/Project managers have different prioritization on what is more important in a candidate?
Do you give whitelabling for this?
Report
Maker
@sagar2agrawal if you are a high volume recruiter we can integrate with your ATS. Otherwise you can use our self-service portal.
All interviews will have features enabled along with bulk export options.
Your jobs are yours and no data entering our system will ever be public.
Scoring works on our internal algorithms, built on millions of profiles and company archetypes.
We do provide a whitelabeling option for enterprise clients.
Report
The Product solve a big challenge in hiring process the user interaction was good. Just wanted to know how do you tailor your interviews as per different job requirements and bring the feel of a real interviewer interviewing you
Report
Maker
@zaid_mahmud We have developed a detailed Role-skill-task knowledge graph. Using this we create a curated set of questions based on the skills, role, and responsibilities mentioned in the JD.
Report
Using agents for initial screening and interviewing targets the right bottleneck. Most hiring funnels clog at resume review and first calls, not at offer. Does the agent adapt its questioning based on role type, or does every candidate get the same flow regardless of seniority?
Report
Maker
@dhiraj_patel5 thanks for your question, Dheeraj! The agent adapts as the conversation goes on and asks contextual questions based on the role.
Report
@saumya_anand1 That's interesting. I have seen so many where the make it compuslory to give followup questions. That totally breaks the purpose of using AI altogether
Report
Maker
@saumya_anand1 @dhiraj_patel5 Yup, that's an important aspect. It is critical for agents to know when to ask follow-up questions and when to move on. Our agents are smartly trained and guard railed to maintain this balance.
Report
we just went through a hiring round and the screening alone took longer than the actual interviews. if this actually catches the candidates you'd shortlist yourself without losing the good ones to slow pipelines thats a real solve
Report
Maker
@tina_chhabra Completely valid and honestly one of the more under-appreciated pain points in hiring. The screening phase is supposed to be a filter, not a bottleneck, but in practice most teams are spending more hours reviewing resumes than they are actually talking to candidates. You end up with slower pipelines, recruiter fatigue, and the best candidates already off the market by the time you get to them.
Here's what we're seeing with clients using SelectPrism's Search & Match on top of their ATS. The AI screens against the Prism Knowledge Graph and ranks the entire pool before a recruiter looks at a single profile. From 854 applications in one recent cycle, the system surfaced 218 recommended candidates - a 74% reduction in volume that recruiters needed to touch at all.
The accuracy piece is what makes it sticky: of the candidates the algorithm marked as top recommendations, 0 were rejected by the recruiter. 100% alignment on the shortlist that actually mattered. Overall, the model hit 94% accuracy across the full funnel.
Report
The 60,000-skill taxonomy built on 60M+ job descriptions is the technically interesting differentiator here. But the hard part isn't the taxonomy size, it's keeping it current. Skills decay and emerge faster than any static dataset can track (a skill that was senior-level two years ago is now table stakes). How does SelectPrism handle taxonomy freshness? Is there a continuous ingestion pipeline from new job postings, or does the graph get periodically retrained on a fixed schedule?
Report
Congrats on the launch! The skill-graph parsing and anti-cheating checks look incredibly robust. Quick question on the ATS integrations: do you currently support GetKnit out of the box, or is that on your upcoming roadmap?
Report
Maker
@abhigyan103 We support out of the box integrations with multiple top ATS. For any ATS that's outside the initial list, we can quickly add those as well.
Report
Thanks for the feedback. We are definitely looking to add additional channels of interviewing which will improve experience for both candidates and companies.