Full-Stack Algorithm Engineer

D2 Intelligence HK Limited; Hong Kong Island; On-site; HKD 550,000 - 1,000,000; Full time

Job summary

D2 Intelligence HK Limited is seeking a versatile Full-Stack Algorithm Engineer to drive end-to-end R&D, optimization, and production delivery for document intelligence, CV, and AI tamper detection across our global products Documind and Checkie.

The role merges algorithm research with full-stack development and direct client engagement, requiring collaboration with international teams and rapid prototyping of AI Agent workflows for production-grade deployment.

Qualifications

• OCR and document information extraction experience.
• AI fake and tamper detection experience.
• Proficiency in PyTorch and TensorFlow.

Responsibilities

• Lead end-to-end R&D, optimization, and production delivery of document intelligence, CV, and AI tamper detection.
• Collaborate with client teams to translate business demands into AI Agent workflows and rapid prototypes.
• Perform on-site system docking, API integration, RAG knowledge base construction, and knowledge graph engineering.
• Abstract reusable modules to upgrade the global product lines.

Skills

OCR recognition

Document information extraction

Image classification

Target detection

AI tamper detection

Seal detection

Signature detection

Font recognition

AI Agent workflows

Python

Java

FastAPI

Flask

SpringBoot

SpringCloud

Docker

Gradio

Streamlit

React

Vue

Cursor

Claude Code

English

Cantonese

Education

Bachelor's degree in Computer Science / AI / Software Engineering

Tools

PyTorch

TensorFlow

TensorRT

ONNX

Knowledge graph

Job description

We are seeking a versatile Full-Stack Algorithm Engineer to join our international product team, focusing on the end-to-end R&D, optimization and production delivery of document intelligence, computer vision (CV), and AI fake tamper detection technologies for our core global products Documind and Checkie. This role combines algorithm research, full-stack engineering development, and client-side project delivery. You will directly engage with global business clients, translate business demands into landing AI Agent systems, and precipitate reusable technical capabilities to drive the iteration and upgrade of standardized international products.

Key responsibilities

Lead the full lifecycle R&D, engineering optimization and production implementation of core capabilities including document intelligent analysis, CV visual recognition and AI fake forgery detection for overseas flagship products Documind and Checkie.

Collaborate closely with front-line client business teams to sort out business procedures and dig into algorithm scenario requirements. Convert vague and unstructured business demands into executable, high-feasibility AI Agent workflow designs, rapidly build and verify AI application prototypes, and iterate steadily to deliver production-grade Agent systems.

Undertake on-site client system docking and technical delivery work, including API integration, RAG knowledge base construction, knowledge graph engineering development, as well as continuous Prompt and Workflow optimization, to ensure stable operation and high-quality delivery of online production systems.

Summarize and abstract general technical capabilities and reusable modules from diversified project practices, feed back to the standardized product system, and empower the continuous iteration and capability upgrading of global product lines.

About you

Bachelor's degree or above in Computer Science, Artificial Intelligence, Software Engineering or other related technical majors.

In-depth practical experience in OCR recognition, document structured information extraction, image classification, target detection, as well as seal, signature and font identification, together with AI fake and tampering detection technologies.

Proficient in mainstream deep learning frameworks including PyTorch and TensorFlow.

Hands-on experience in fine-tuning vision large models (VLM/Vision-LLM) and open-source/commercial document intelligent models, as well as inference acceleration optimization via TensorRT/ONNX and industrial engineering integration.

Capable of building interpretable risk feature extraction pipelines for complex business documents such as commercial invoices, contracts, and KYC/KYB account opening files.

Master Python/Java programming languages, proficient in mainstream development frameworks including FastAPI, Flask, SpringBoot and SpringCloud.

Able to independently complete high-performance model inference API encapsulation, standard RESTful interface design and Docker containerized deployment.

Skilled in using Gradio, Streamlit and other rapid development tools to build algorithm demonstration platforms. Possess basic React/Vue front-end development capabilities to support product demonstration and interactive iteration.

Rich practical experience in leveraging AI coding tools such as Cursor and Claude Code to efficiently complete independent project development and product construction.

Fluent in both English and Cantonese with excellent bilingual reading and writing skills, capable of independently compiling professional English technical documents and customized client solution proposals.