Looking for a job

Arijit De

@arijitde

Message

Machine Learning Engineer specializing in multi-modal RAG, multi-agent AI, and computer vision.

India

Message

What I'm looking for

I’m looking to build production-ready multi-agent and multi-modal ML systems—RAG pipelines, computer vision, and MLOps—with strong reliability, measurable evaluation, and scalable cloud deployment.

I’m a Machine Learning Engineer with a PhD and 6 years of industry experience spanning computer vision, multi-agent AI systems, and MLOps. I’ve designed and built scalable, robust multi-modal RAG pipelines and 3D/2D image segmentation and object detection solutions across construction, healthcare, and automotive domains.

Recently, I led the development of a multi-agent platform for construction “Autonomous Project Management & Scheduling,” powering intelligent project information retrieval for 100 concurrent users / 3000 agent runs per day and improving business efficiency by 50%. I also hardened workflows with failure-mode handling to reduce task failure rate by 60% and sustain 99.4% uptime, while improving evaluation scores by 75% using CrewAI Tracing and DeepEval. Earlier roles included delivering 3D computer vision that improved diagnosis speed by 40% and achieved 88.68% accuracy, and optimizing model deployment with ONNX/Docker on GCP to reduce inference latency by 65%.

Experience

Work history, roles, and key accomplishments

Current

Machine Learning Engineer

Current

Mvizn

Jan 2024 - Present (2 years 6 months)

Led a multi-agent construction platform (CrewAI, Python, FastAPI), boosting efficiency 50%. Hardened workflows for 99.4% uptime and -60% failures. Built multi-modal RAG (Pinecone, 10M vectors) with 92% accuracy, <600ms latency. Fine-tuned Gemini LLMs via LoRA, raising validity to 98% and cutting costs 30%. Integrated DeepEval tracing, lifting relevancy scores 75%.

Crew AI Python fastAPI Multi agent orchestration Pinecone Hybrid Search Reranking Observability DeepEval Retries Timeouts cloud deployment

Machine Learning Engineer

Synerjix

Jun 2025 - Dec 2025 (6 months)

Built a Smart Moving Assistant that estimates shifting cost of items from room photos, improving stakeholder business efficiency by 50%, adopted by 3500 users / 10000 estimates processed.
Deployed DinoV3-based object detection at 65.3 mAP, serving inference at 10ms via Python + FastAPI, MySQL, and AWS (SageMaker, EC2, EKS) with CI/CD and Docker.

Python fastAPI Object Detection MySQL AWS Sagemaker AWS EC2 CI CD Docker

Computer Vision Engineer

Institute of Neurosciences

Jan 2022 - Dec 2023 (1 year 11 months)

Engineered 3D Computer Vision solutions (PyTorch, OpenCV) for neurological analysis, improving diagnosis speed 40% and achieving 88.68% accuracy. Scaled MLOps using MLflow, cutting model iteration time by 50%. Optimized models with ONNX and deployed on GCP via Docker, reducing inference latency by 65%.

PyTorch Python OpenCV NumPy 3D Computer Vision MLFlow Model Registry Reproducible Training Onnx Docker GCP IAM Cloud SQL

Machine Learning Engineer

Mercedes-Benz Research and Development India

Aug 2018 - Sep 2019 (1 year 1 month)

Developed and optimized a computer vision pipeline for Vulnerable Road User (VRU) detection by fine-tuning YOLOv3 on large-scale annotated datasets. Designed and deployed AWS workflows for training/evaluation and built automated Python-based QA and evaluation frameworks.

Computer Vision VRU Detection YOLOv3 Model Training Model Evaluation AWS Python Quality Assurance Annotation Throughput Data Quality

Systems Engineer

Tata Consultancy Services

Aug 2014 - Apr 2015 (8 months)

Maintained high-availability Java/MySQL backend systems for TCS’s internal website serving 100,000+ users. Streamlined the feature release cycle to reduce deployment time.

Java MySQL High Availability Backend Systems Release Cycles

Education

Degrees, certifications, and relevant coursework

Jadavpur University

Doctor of Philosophy (PhD), Computer Science

2020 - 2024

Grade: 9

Activities and societies: Guest Faculty at Programming Lab

Researched AI/ML-based detection, classification, segmentation and prediction algorithms using Deep Learning and Computer Vision for neurological disorders via radiological and histopathological imaging.