Python Software Developer - Computer Vision & Document Intelligence

# Python Software Developer – Computer Vision & Document IntelligenceFi Systems ·, CABack to resultsFIPython Software Developer – Computer Vision & Document IntelligenceFi Systems, CA Posted November 4, 2026 Via LinkedIn### Full Job DescriptionAbout Fi SystemsFi Systems develops AI-powered applications and assistants that help engineering and procurement teams work more efficiently.We’re looking for a software developer to help build a platform that turns complex HVAC and engineering drawings into structured, usable data. The software analyzes PDFs, identifies symbols and technical information, and presents the results for users to review.If you enjoy figuring out why an algorithm works on one drawing and fails on another—and turning that investigation into a reliable improvement—this role could be a good fit.About the RoleYour primary focus will be developing and improving our Python extraction algorithms. You’ll combine PDF text and vector geometry, computer vision, OCR, and AI vision services to interpret drawings with different layouts, scales, and levels of quality.You’ll also contribute to the C#/.NET backend that manages documents, coordinates processing, stores results, and exposes APIs.Your strongest experience should be in Python and practical extraction or image-processing work. Experience with C#/.NET is valuable, but we also welcome strong Python developers who have built production backend services in another language and are ready to learn our .NET stack.This is a hands-on software development role, with responsibility for taking improvements from investigation through implementation, testing, and validation.What You’ll Work On

• Extract information from technical drawings: Detect legends and symbols, match symbols across drawings, and associate them with descriptions, equipment tags, dimensions, and airflow information.
• Improve accuracy across varied documents: Investigate missed detections, incorrect matches, duplicates, rotations, ambiguous annotations, and differences between native and scanned PDFs.
• Combine complementary techniques: Use native PDF content, geometric reasoning, image processing, OCR, computer vision, and AI services to produce reliable results.
• Measure and validate improvements: Build evaluation datasets and regression tests, track precision and recall, and balance accuracy with processing speed, memory usage, and service costs.
• Connect extraction to the application: Develop backend APIs, database models, and asynchronous workflows, including service integration, retries, and error handling.
• Build maintainable software: Participate in code reviews, document technical decisions and limitations, and collaborate with developers and engineering domain specialists.What You Bring
• Strong Python development skills, typically gained through 3–5 years of relevant experience or equivalent substantial software projects.
• Practical experience with computer vision, image processing, OCR, or document extraction, using OpenCV, NumPy, or comparable tools.
• Confidence working with coordinates, bounding boxes, transformations, matching scores, and other geometric or spatial concepts.
• Experience building and maintaining backend services and REST APIs, ideally with C#/.NET, or a strong foundation in another backend stack and willingness to learn.
• Experience with relational databases, automated testing, Git, and systematic debugging.
• The ability to investigate an unfamiliar problem, explain your approach, and carry a solution through implementation and validation.A degree in computer science, software engineering, computer engineering, or a related discipline is welcome. Equivalent professional experience is equally relevant.Experience That Would HelpYou don’t need experience with every tool or topic below. We’re particularly interested in candidates who bring some of the following:
• PDF extraction.
• Engineering drawings, schematics, CAD-generated documents, or HVAC terminology.
• Template matching, segmentation, shape comparison, or linking detected objects to nearby text.
• Multimodal AI, structured model outputs, and validation of AI-generated results.
• ASP.NET Core, Entity Framework Core, and SQL Server.
• Azure AI Document Intelligence, Azure AI Foundry, object storage, message queues, or Docker.Why Join Us?
• Solve a tangible engineering problem. Help turn complex drawings into information people can review and use.
• Influence the core technology. Contribute directly to extraction methods, evaluation approaches, and backend design.
• See measurable results. Evaluate your work through improvements in accuracy, speed, reliability, and user experience.
• Develop across disciplines. Work at the intersection of Python, computer vision, document intelligence, and backend software development.
• Contribute your judgment. Propose improvements and collaborate with a team that values curiosity, ownership, and maintainable code.Work Arrangement and BenefitsThis role requires 2 days per week at our Boucherville office, with remote work for the remainder of the week.Benefits include:
• Extended health, dental, and vision care.
• Life and disability insurance.
• Paid time off.
• On-site parking and casual dress.Interested?Apply with your résumé. If possible, include a brief description of a relevant project: the problem you tackled, what you personally built, and how you evaluated the results.Examples involving PDFs, OCR, image processing, computer vision, or spatial algorithms are especially useful. A short description is sufficient if your work is confidential.#J-18808-Ljbffr