PulseHR - AI-Powered HR Feedback & Performance Review Analytics
Talent Management teams and line managers repeatedly create performance-review questionnaires, collect manager-to-employee and employee-to-manager feedback, remind people to complete forms, and manually summarize results. This repetitive work consumes HR time and makes it difficult to identify the most important issues across teams. PulseHR transforms a short HR brief into an editable survey, provides a shareable response link, calculates descriptive statistics, and uses AI to prioritize key findings and suggest follow-up actions. It helps HR teams make better-informed decisions without replacing human judgment in employee evaluations. Problem discovery was informed by an informal, non-confidential account from a Talent Management intern.
Preliminary score
- Value for the user16.0 / 25
A specific problem and a clear outcome from solving it.
- Prototype and use of AI19.5 / 30
A working core scenario, and what AI actually contributes to the result.
- Quality testing4.5 / 20
Tests, examples of failures, and comparison with the current approach.
- Feasibility7.5 / 15
Data requirements, running costs and a clear next step.
- Originality4.0 / 10
A meaningful difference from common solutions.
What the judges wrote
- Claude Opus 5.5Preliminary48/100
- Strengths
- Your screenshots show a coherent end-to-end flow: an HR brief becomes an editable survey, responses are collected, per-question statistics are computed, and the AI produces a prioritized table. Each row in that table pairs an observation with a supporting data point and a concrete next step. Grounding the AI report in computed aggregates, rather than raw text alone, is a sensible design choice. Your privacy touches stand out for HR data: names and emails are kept out of the prompt, and HR must explicitly consent before open-text comments are sent to the model. The human-in-the-loop review step before publishing is clearly built into the workflow.
- To improve
- Your deck says PulseHR was 'rigorously tested' on question relevance, statistical accuracy and edge cases, but no test cases, numbers or failures are shown. The only issue reported is an API key misconfiguration. Run a small, systematic evaluation, for example 10 survey briefs rated for question quality and AI priorities checked against a human HR reviewer, and time it against the manual consolidation it replaces. The track asks you to name an adoption metric, such as hours saved per review cycle or time-to-report, so state one and estimate it. Also add running costs per survey or report on Groq, and explain how you differ from existing tools such as Culture Amp or Qualtrics.
- GPT-6 AstraPreliminary55/100
- Strengths
- Your project addresses a concrete HR workload through an integrated survey-to-report workflow, with meaningful AI roles in question generation and prioritization. Your screenshots show a five-response example with numerical summaries and an AI report connecting observations to supporting data and next steps. Human review, synthetic-only demonstration data, and explicit disclosure of missing enterprise controls establish sensible MVP boundaries.
- To improve
- Your testing evidence should distinguish completed checks from planned work: the deck claims verified statistical accuracy, while the submission lists that validation as a next step. Provide reproducible expected-versus-actual results for statistics, small samples, and unsupported AI conclusions, alongside a complete brief-to-report example. Measure HR time saved against a forms-and-spreadsheet workflow with an HR user. Add per-survey operating costs and a concrete plan for access controls, production persistence, and protecting identifiable free-text feedback.
The team
Nihatologiya Ismayilzada
data
Vahid MammadovLead
product
Agil Novruzov
developer