Background Texture Mobile
PATHIXO INNOVATION & CLIENT SHOWCASE

Engineering Breakthroughs in AI, Web & Systems

From our in-house flagship desktop assistant Vakxo to enterprise client platforms, fine-tuned spiritual AI models, and data science research.

Flagship In-House Product

Meet Vakxo

Your Live Autonomous Desktop AI Buddy.

Talks, listens, and takes control of your PC. Vakxo runs natively on your machine, pairing with you as an autonomous co-pilot to write code, navigate IDEs, run terminal commands, and execute system workflows without lifting a finger.

Q3 2026
Target Release
Native
Desktop Execution
< 250ms
Voice Latency
Windows Preview
macOS
Linux
vakxo-agent :: active
LISTENING

Talks & Listens Bi-Directionally

Ultra low-latency speech processing allows natural back-and-forth verbal dialogue while you work.

Codes With You Locally

Executes CLI commands, refactors source code, and inspects files in your real desktop environment.

Autonomous System Computer Use

Navigates browser windows, launches software, and handles repetitive desktop administration securely.

Electron / RustNative OS APIsSpeech-to-Speech LLMComputer VisionTypeScript
Client Case Study • Healthcare & Telemedicine

Bisvaran Healthcare Platform

Bisvaran Healthcare Platform UI Preview
Click to Inspect UI
bisvaran.com
In-Home ICU Nursing
Doctor Home Checkups
Bedside Attendants
Blood Sample Collection
Elderly Companionship
Specialized Therapies

Professional Home Healthcare & Telemedicine Portal

Client: Bisvaran Healthcare Pvt. Ltd.

Bisvaran required a high-conversion, empathetic, and rapid triage web platform to bridge the gap between hospital-level medical requirements and comfortable at-home healing. Pathixo engineered a fast, accessible portal equipped with one-click doctor consultations, home nursing scheduling, bedside attendant requests, and direct laboratory blood sampling dispatch.

In-Home ICU & Specialized Nursing: Seamless care requests for post-operative recovery, chronic illness, and palliative care.
Doctor Consultation & Direct Home Visits: Instant video appointments and on-site checkups with verified medical practitioners.
Dedicated Bedside Attendants: 12-hour and 24-hour caregiver booking for hospital stays and post-discharge mobility.
Safe In-Home Blood Sampling: Hygienic lab sample collection direct from patients' doorsteps.
Next.jsTypeScriptTailwind CSSFramer MotionTelehealth EngineSEO Architecture
Explore Portal
Fine-Tuned LLM • Spiritual AI Companion

My Kanha — Converse with Krishna

An AI companion trained on the complete sacred verses and philosophy of the Bhagavad Gita, delivering timeless wisdom and inner peace.

Bhagavad Gita 18 Chapters • 700 Shlokas

Timeless Counsel for the Modern Seeker

My Kanha brings ancient spiritual heritage to modern seekers. Fine-tuned extensively on all 18 chapters and 700 sacred shlokas of the Bhagavad Gita, the AI serves as a compassionate guide for life's toughest moments—unpacking Dharma, Karma, fear, devotion, and purpose. Built as a Progressive Web App (PWA) with bespoke spiritual themes.

Bhagavad Gita Contextual Intelligence: Provides profound philosophical and spiritual perspectives rooted in authentic Vedic scriptures.
Empathetic Conversational Persona: Replies in the soothing, compassionate, and elevated tone of Lord Krishna/Kanha.
Progressive Web App (PWA): Installable directly on Android, iOS, and desktop for offline reflection and instant access.
Custom Spiritual Theme Engine: Tailored color palette customization (Sacred Saffron, Golden Ochre, Dark Mode toggle).
Multilingual Dialogue: Seekers can inquire and receive spiritual insights in both English and Hindi.
Fine-Tuned LLMNext.js / ReactPWA (Service Worker)Custom Theming EngineTailwind CSS
700 Shlokas
Scriptural Foundation
18 Chapters
Chapters Covered
Installable PWA
App Format
Converse with Krishna — Welcome Screen

Converse with Krishna — Welcome Screen

Interactive Dialogue & Live Spiritual Counseling

Interactive Dialogue & Live Spiritual Counseling

Settings & Spiritual Theme Engine

Settings & Spiritual Theme Engine

Click any screenshot to view full resolution details & theme customizer

Machine Learning & Data Intelligence

Data Science Research & Predictive Analytics

Quantitative modeling, multi-variable regression manifolds, and unsupervised clustering applied to market intelligence and smart city infrastructure.

Machine Learning Valuation Engine

Mobile Market Intelligence & Price Predictor

By cleaning and modeling extensive smartphone market datasets, this project uncovers hidden pricing structures across major global manufacturers (Samsung, Apple, Realme, OPPO, Nokia, Xiaomi). Incorporates a proprietary Value Index to identify true performance-per-rupee value, and deploys a live Streamlit dashboard for interactive price prediction.

3D Multi-Variable Regression: Discovered non-linear relationship surfaces mapping RAM (GB), Storage (GB), and Market Price (₹).
Brand Value Index Analysis: Quantified consumer performance return against brand markups across global brands.
Exploratory Distribution Modeling: Detailed statistical histograms of prices, memory configurations, and user sentiment ratings.
Streamlit Inference Dashboard: Interactive web application allowing users to input phone specifications and predict fair market valuation.
Outlier Detection & Data Cleansing: Robust preprocessing pipeline ensuring zero skews from promotional anomalies.
PythonScikit-LearnPandasNumPyPlotly 3DStreamlitRegression Modeling
3D RAM, Storage & Price Manifold

3D RAM, Storage & Price Manifold

Value Index by Brand (Higher is Better)

Value Index by Brand (Higher is Better)

Top 10 Brands by Market Share

Top 10 Brands by Market Share

Distribution of Mobile Phone Prices

Distribution of Mobile Phone Prices

Top 15 Mobile Phone Brands by Model Count

Top 15 Mobile Phone Brands by Model Count

Multi-Feature Distribution Quadrants

Multi-Feature Distribution Quadrants

Click any chart to zoom into 3D regression manifolds and brand distributions

Smart City IoT & Clustering Analysis

Smart City Parking Analytics & ML Clustering

Urban parking bottlenecks contribute significantly to urban traffic delays and fuel wastage. In this project, we cleaned, normalized, and modeled extensive municipal parking data, mapping occupancy rates across hours and days, and performing K-Means clustering to classify parking lots based on utilization elasticities.

K-Means Lot Clustering: Grouped urban facilities into distinct congestion archetypes based on mean occupancy and capacity.
Diurnal Time-Series Profiling: Visualized hourly occupancy shifts from 08:00 to 17:00, mapping peak commercial influx hours.
Day-of-Week Trend Analysis: Measured volume oscillations across Tuesday–Sunday with significant mid-week peaks.
Facility Utilization Benchmarking: Pinpointed chronically strained lots (>75% occupancy) vs underutilized regional spaces.
PythonK-Means ClusteringSeabornMatplotlibPandasTime-Series Modeling
Mean Occupancy Rate by Facility ID

Mean Occupancy Rate by Facility ID

Occupancy Rate vs Time of Day

Occupancy Rate vs Time of Day

Weekly Volume Influx

Weekly Volume Influx

Click any chart to zoom into occupancy progressions and municipal facility clusters

Have a Vision for an AI or Web Product?

Whether building an autonomous desktop application, high-converting digital service, or custom fine-tuned LLM, our engineering team brings it to life.