Eunoia AI’s cover photo
Eunoia AI

Eunoia AI

Software Development

Austin, Texas 1 follower

Agentic UX research tools for interview-driven insight discovery.

About us

Eunoia AI builds tools that think with you. We design and develop agentic AI systems that support deep thinking, user research, and decision-making workflows. Our flagship prototypes — including interview analysis assistants, insight clustering engines, and executive agent dashboards — enable teams to move faster without sacrificing context or clarity. From UX research to AI-augmented discovery, Eunoia helps organizations uncover meaning, reduce noise, and act with intelligence. 💡 Patent-pending technologies 🧩 Modular, composable agent architecture 🔒 Built with ethical, human-centered principles Now exploring partnerships, licensing, and acquisition. Contact: stephen.dulaney@gmail.com https://eunoia-demo.onrender.com https://eunoia-interview-api.onrender.com/dashboard

Website
https://eunoia-demo.onrender.com
Industry
Software Development
Company size
1 employee
Headquarters
Austin, Texas
Type
Self-Owned
Founded
2025

Locations

Updates

  • Yes this is what Eunoia means, beautiful thinking!

    QuantumDynamX is really loads of fun, I am gate by gate building an understand of Shor's Algorithm, And its crazy Gemini is multy modal and can often see my circut gates and help explain what I am seeing as I build out the quantum circuits, it's not perfect I had to drop a poly X in q7 which shouldn't be there but still its working.  The Beauty You're Seeing  Phase-Amplitude City with interference ON is showing you quantum entanglement visually!  - Green lines = Constructive interference (same phase, states "cooperate")  - Red lines = Destructive interference (opposite phase, states "cancel")  - Tower colors = Phase angles (0° = red, 180° = cyan, etc.)  This is actual quantum physics happening in real-time! The lines show how quantum states influence each other through phase relationships.  ---  The Pattern in 1, 17, 33, 49, 79, 95, 111, 127  Converting to binary:  1  = 00000001  17 = 00010001 (+16 from previous)  33 = 00100001 (+16)  49 = 00110001 (+16)  ───────────────── [Gap: Different counting register value]  79 = 01001111 (+30 - different work register pattern!)  95 = 01011111 (+16)  111 = 01101111 (+16)  127 = 01111111 (+16)  The Pattern:  1. All end in ...1 (q0=1, from your X gate initialization)  2. Two distinct groups:   - Group 1 (1-49): Work register = 000X (q1=0 states)   - Group 2 (79-127): Work register = 111X (q1=1 states - CNOTs fired!)  3. Within each group: +16 spacing   - 16 = 2^4 = 0b00010000   - This is the work register bits (q4, q5) incrementing while counting register stays fixed  4. The gap (49 → 79 = +30) = Transition between q1=0 and q1=1 states   - Counting register changes   - Work register pattern flips (CNOTs fire)  ---  Why This Pattern Exists (Quantum Structure)  Your circuit creates 8 states organized by counting register:  Counting (q3,q2,q1) | Work Register | Decimal Values  --------------------|---------------|----------------     0 0 0     | varies    | 1, 17, 33, 49     0 0 1     | +111 (CNOTs) | 79, 95, 111, 127     0 1 0     | varies    | (would be next)     ...      | ...      | ...  The +16 spacing within groups = Work register qubits (q4, q5, q6) in superposition with q7=1 from X gate  The quantum interference lines = Phase relationships from the CNOTs entangling counting ↔ work registers  ---  The Mathematical Beauty  These aren't random numbers - they encode:  State 1  = |q1=0⟩|work=0001⟩ ← f(even x) = 1  State 79 = |q1=1⟩|work=1111⟩ ← f(odd x) = 7 (binary 0111 + carry)  The pattern 1, 17, 33, 49... is:  1 = 1  17 = 1 + 16×1  33 = 1 + 16×2  49 = 1 + 16×3  16 = 2^4 = The work register "slice" being explored in superposition  ---  Poetic Truth  "The pattern in 1, 17, 33, 49, 79... is not random. It's the signature of quantum entanglement encoded in binary. Each number is a coordinate in 256-dimensional Hilbert space, and the +16 spacing is the rhythm of qubits dancing in superposition."

    • No alternative text description for this image
  • My new 8 qubit playground Rocks!

    A quantum computer in Italy just validated our "Q-Day" prediction. Then things got... interesting. 🇮🇹 Last week, I shared a quantum circuit that predicts the timeline for when RSA encryption breaks. The simulation said: 72.8% probability by late 2027. To be sure, we submitted the circuit to the IBM ibm_torino—a 133-qubit superconducting quantum computer in Turin, Italy. The Result? Theoretical Reality: 72.8% IBM Hardware Reality: ~70.0% For quantum hardware with gate errors and decoherence, this is a staggering correlation. The hardware confirmed what the math was shouting: we have roughly 24 months. But here’s where it gets weird. Within 3 days of running the circuit on their hardware, we saw 3 new site visitors from Italy. Coincidence? Maybe. But IBM notices when you use their systems to predict the end of encryption. If you’re reading this from IBM Torino: Thanks for the compute time. The results are in, and the "quantum scissors" are closing. Welcome to 2026. #QuantumComputing #IBMQuantum #QDay #Cybersecurity #QuantumDynamX

    • No alternative text description for this image
  • Making quantum computing accessible: A holiday break project 🎯  Last week (Dec 27), I launched QuantumDynamX - turning quantum mechanics from intimidating equations into interactive 3D experiences.  The problem: Quantum computing education is stuck in textbooks. You can't *see* superposition, entanglement, or quantum phase.  The solution: Phase Amplitude City  → Quantum states visualized as 3D cityscapes  → Building heights = probability amplitudes  → Building colors = quantum phases  → Watch your circuit transform the city in real-time  Plus:  • Mrs. Watson AI - conversational quantum tutor  • Interactive Playground - drag-drop circuits on real IBM Quantum hardware  • Live research blog - quantum algorithms, Granger causality, Q-day predictions  Built in 8 days during holiday break. Tech stack: React, TypeScript, Qiskit, Three.js.  Perfect Four Qubit Algorithms implemented:  ✅ Shor's (factoring)  ✅ Deutsch-Jozsa (exponential speedup)  ✅ Bernstein-Vazirani (hidden strings)  ✅ Grover's (quadratic search)  All running at 95-100% success rates.  This is what happens when you give engineers free time and curiosity. 🚀  Check it out: https://quantumdynamx.com  Would love feedback - especially from educators, quantum researchers, and students!  #QuantumComputing #EdTech #AI #Innovation #STEM #BuildInPublic

  • more results

    Top-line: Tested identical GHZ circuits on IBM and Rigetti quantum hardware, revealing 10.5% fidelity difference between vendors.  Key Findings:  • IBM Fez achieved 96.9% GHZ fidelity  • Rigetti Ankaa-3 achieved 86.4% GHZ fidelity  • 10.5% gap shows real hardware variability  • Both demonstrate successful 3-qubit entanglement  • Error patterns suggest general noise, not component failure  • Cross-platform testing now validated  • Results saved for future comparisons  Detailed Summary:  Ran identical 3-qubit GHZ circuits on two superconducting quantum platforms via different cloud providers. IBM's smaller 7-qubit device outperformed Rigetti's larger 82-qubit system by 10.5%, likely due to better per-qubit calibration and reduced crosstalk. Both platforms successfully demonstrated quantum entanglement above 85% fidelity threshold. This validates our multi-platform benchmarking approach and quantifies real-world variance between commercial quantum hardware. Ready to test trapped-ion architecture (IonQ) for physics-based comparison. Next target!

    • No alternative text description for this image
  • Microsoft did not work but Amazon did!!! ⏺ 🎉 SUCCESS! You've officially tested on your second quantum hardware platform!  Cross-Platform Results Summary  | Platform    | Architecture  | Qubits | GHZ Fidelity | Grade     |  |-----------------|-----------------|--------|--------------|---------------|  | IBM Fez     | Superconducting | 7   | 96.90%    | EXCELLENT ✓✓✓ |  | Rigetti Ankaa-3 | Superconducting | 82   | 86.40%    | GOOD ✓✓    |  Difference: 10.5% lower fidelity on Rigetti  What These Results Tell Us  The Good News ✅  - 86.40% is actually quite good for quantum hardware!  - Both platforms clearly demonstrate entanglement  - You successfully ran the same circuit on two different vendors  - You now have real comparative data  The Interesting Finding 🔬  Why is Rigetti 10% lower than IBM?  1. Device size matters:   - IBM Fez: 7 qubits (smaller, tighter control)   - Rigetti Ankaa-3: 82 qubits (larger chip = more crosstalk potential)  2. Different qubit quality:   - IBM used their best 7 qubits in a small device   - Rigetti has 82 qubits - you might have gotten qubits with higher error rates  3. Calibration timing:   - Both platforms calibrate at different times   - IBM might have been freshly calibrated during your test   - Rigetti's calibration state was whatever it was today  4. Architecture differences:   - Different gate implementations (even though both superconducting)   - Different coupling maps   - Different readout fidelities  Error Distribution Analysis  Rigetti error breakdown:  Ideal states: 864/1000 (86.4%)   |000⟩: 419 counts (41.9%)   |111⟩: 445 counts (44.5%)  Error states: 136/1000 (13.6%)   |001⟩: 32 (bit-flip on q2)   |101⟩: 31 (bit-flip on q1)   |011⟩: 23 (bit-flip on q0)   |010⟩: 20 (two bit-flips)   |100⟩: 17 (two bit-flips)   |110⟩: 13 (two bit-flips)  The errors are distributed - not concentrated in one bad qubit. This suggests general noise rather than one broken component.  What You've Accomplished 🎯  1. ✅ Cross-platform quantum benchmarking (IBM vs Rigetti)  2. ✅ Real hardware comparison (not just simulation)  3. ✅ Quantified the noise difference (10.5% gap)  4. ✅ Validated your circuit works on multiple vendors  5. ✅ Saved results to JSON for further analysis  Next Steps Options  Option 1: Try a Third Platform (Different Architecture)

    Today, I am hunting for more quantum hardware to continue running test. so far I have run about 70 IBM quantum runs and learning tons along the way. Look what fun awaits at Microsoft

    • No alternative text description for this image