Computer Vision & OCR Engineer
First seen by Alion on Oct 4, 2026.
Computer Vision & OCR Engineer (Supply Chain Document AI)
Location : Gurgaon / Remote (Hybrid / Flexible Work Policy)
Employment Type : Full-Time
Experience Required : 3 - 6 Years
About Us & Our Mission :
Global supply chain logistics generate an overwhelming volume of unstructured physical and digital paperwork daily - including bills of lading, customs declarations, shipping manifests, invoice receipts, and multi-lingual transport waybills. Transforming these unstructured physical documents into structured, machine-readable data feeds is essential for automating modern logistics analytics.
Our engineering culture values advanced computer vision, deep learning model optimization, and high-accuracy document intelligence pipelines. If you want to apply cutting-edge multimodal AI and optical character recognition to automate global supply chain paper trails, you will thrive here.
The Role & Impact :
We are seeking an innovative and meticulous Computer Vision & OCR Engineer to build, fine-tune, and scale our document intelligence and visual parsing pipelines. In this role, you will develop state-of-the-art OCR and layout-analysis models capable of extracting structured data from complex, low-quality, multi-lingual supply chain documents with near-perfect accuracy. You will work hand-in-hand with our AI and backend squads to integrate these vision models into production-grade ingestion workflows.
Key Responsibilities & Daily Expectations :
- Model Development & Fine-Tuning : Build, fine-tune, and optimize deep learning models for text detection, text recognition (OCR), and document layout analysis (using PaddleOCR, TrOCR, Donut, or LayoutLM).
- Document Parsing Pipelines : Design and scale end-to-end computer vision pipelines that ingest raw images/PDFs, correct skews, perform table extraction, and output structured JSON.
- Inference Optimization : Optimize model weights, quantization (INT8/FP16), and serving runtimes (TensorRT, ONNX Runtime) to achieve low-latency inference on cloud GPUs.
- Data Annotation & Augmentation : Oversee dataset curation, synthetic data generation, and active learning loops to continuously improve model accuracy on edge-case documents.
- System Integration : Integrate vision microservices seamlessly with Python/FastAPI backend workflows and vector search databases.
What We Are Looking For (Requirements) :
- Experience : 3 to 6 years of professional machine learning or computer vision engineering experience with a strong focus on OCR, document AI, or image processing.
- Technical Mastery : Deep proficiency in Python, PyTorch or TensorFlow, OpenCV, and modern transformer-based vision-language architectures.
- Production Deployment : Hands-on experience containerizing models with Docker and deploying high-throughput inference endpoints using FastAPI, Triton Inference Server, or TorchServe.
- Problem-Solving Ability : Proven track record of handling noisy, real-world visual data, skewed scans, and multi-lingual character sets.
- Education : Degree in Computer Science, Artificial Intelligence, Electronics, or equivalent practical research background.
What We Offer & How to Apply :
- Competitive Compensation : Attractive salary package (20 - 34 LPA) with performance bonuses and equity options.
- Advanced AI Stack : Hands-on experience working with state-of-the-art vision models and high-end GPU infrastructure.
- Growth & Wellness : Comprehensive health insurance coverage and annual learning stipend.
Skills
Computer Vision, Optical Character Recognition, Deep Learning, Data Annotation, Python, Image Processing, Tensorflow