AI products
We design and build AI-powered products end to end — research and learning tools, scientific desktop applications, and intelligent assistants.
ignite( ideas ) → working AI
We build AI that works.
AIGNITE is an AI development company based in Vietnam. We design intelligent products and apply AI to real problems — from drug discovery and climate tech to agriculture, education, and public services.
biomedical
climate
agriculture
education
civic services
what we do
From concept to working software
Applied AI & ML
We take AI beyond demos: custom-trained models, RAG systems, computer vision, and scientific computing applied to real domain problems.
Full-stack delivery
From desktop and mobile apps to edge deployments and offline-first PWAs — we design, build, and ship complete systems end to end.
capabilities
Demonstrated across shipped projects
LLM product engineering
Multi-provider integration behind an abstraction layer, RAG, schema-constrained output, real-time audio, prompt-cache cost engineering, and prompt-injection defense-in-depth.
Machine learning & modeling
PyTorch multi-task networks, gradient boosting, time-series forecasting, interpretability with SHAP, and small-data rigor with honest uncertainty.
Computer vision
YOLOv8 detection and segmentation, edge + cloud inference splits, classical CV, and real-time browser detection and DSP at interactive frame rates.
Scientific computing
Molecular docking, ADMET screening, fragment-based de novo design, adsorption-cycle simulation, techno-economic optimization, and physics + ML hybrids.
Full-stack engineering
Tauri desktop apps with embedded scientific Python, React web, React Native mobile, offline-first PWAs, real-time WebSocket, and internationalization.
Embedded & robotics
Microcontroller firmware with fail-safes, RTK GNSS positioning, checksummed UART protocols, sensor integration, and on-device inference.
Cloud, edge & data
Serverless containers on Cloud Run, edge serverless on Cloudflare Workers with geo-gating and rate limiting, size-optimized Docker, and scheduled serverless crawlers.
selected work
Applied AI, in the field
Portfolio shown at sector level — client details stay confidential.
biomedical · pharma r&d
Drug-discovery desktop platform
An end-to-end computational drug-discovery platform pairing multi-task deep-learning resistance prediction with molecular docking, ADMET screening, and fragment-based de novo design — packaged as a cross-platform desktop app with an embedded scientific Python runtime.
multi-task ANNdocking
ADMETtauri desktop
climate tech · ccu
Carbon-capture decision support
A hybrid physics + machine-learning platform for carbon capture and utilization: ML models over experimental catalysis data, an adsorption-cycle simulator, and a techno-economic optimizer — delivered as a bilingual cloud web app with emissions forecasting.
xgboostforecasting
simulationcloud run
precision agriculture
Field-robot vision platform
A sensor-equipped field robot with RTK positioning streaming crop imagery to cloud-hosted YOLOv8 inference, with a real-time dashboard, an advisory engine, and mobile push alerts — from firmware to cloud.
yolov8rtk gnss
firmwarereact native
education · music
AI music-education app
Schema-constrained LLM lesson and quiz generation combined with real-time browser audio: live pitch detection for vocal coaching, streaming audio to multimodal models, and an interactive multi-instrument simulator.
multimodal llmpitch detection
ttsreact spa
civic services · client work
Offline-first form-assistance PWA
An offline-first PWA that guides users through administrative forms, pairing a zero-cost deterministic instruction layer with a geo-gated, rate-limited LLM chat backend on edge serverless infrastructure — provider-swappable models, lexical RAG that cuts prompt tokens 4–10×, and defense-in-depth output filtering.
offline-first pwacloudflare workers
ragcost engineering
prompt-injection defense
company
AI + ignite — igniting ideas into built, working AI
AIGNITE is an AI-focused software development company based in Hải Phòng, Vietnam. We believe the value of AI is realized only when it ships — as a product someone uses, a system someone relies on, a tool that quietly works.
That's why everything we show here is evidence, not promise: products we run, systems we've built, and domains we've shipped into. Concrete capabilities over hype.
ignite( your idea )
Have a problem AI could solve?
Tell us what you're trying to build — we'll tell you honestly whether AI is the right tool.