An image of a 3D stack, labeled 'Encoder', surrounded by tech-related icons, against a dark background **Original Explanat...

Applied Machine Learning, Mathematics, Software Engineering, and Real World Solutions

Mridul Sarkar

Exploring tower-based and universal encoders for advancing multimodal AI toward artificial general intelligence. Building scalable, production-ready systems across diverse data types and real-world problems.

Expertise in Multimodal AI and Machine Learning

Mridul Sarkar specializes in developing scalable machine learning systems that integrate vision, audio, text, and time series data. His focus is on creating unified models, tower-based encoders, and universal architectures that push the boundaries toward artificial general intelligence.

An illustrated view of a server room with a network of nodes linking together in the foreground, representing the concept ...

Always building and learning!

A man in a white shirt looks at a whiteboard with a black line drawing of a speech bubble, a heart, and a megaphone connec...

Mridul Sarkar’s Blog

Discover insights into Mridul’s recent work.

  • Languages: Python, Typescript, SQL, Bash
  • Cloud: AWS Bedrodck, SageMaker, EC2, S3, Lambda, CloudFront, DynamoDB, Route53, CloudWatch, CloudFormation, IAM
  • Frameworks & Tools: FastAPI, React, Celery, MongoDB, OpenSearch/ElasticSearch, MATlab, Triton Inference Server
  • ML & Quantum: HuggingFace, HG Transformers, Qiskit, Pennylane, PyTorch, Scikit-learn, NLP, Tensor Networks, DSPy
  • Testing & DevOps: Unit & Integration Testing, TDD, Git, Docker, CI/CD Pipelines, IaC
  • Business Development: EQ-driven leadership, clear comms, business-case storytelling, and tech-to-strategy translation

Visit us

Chicago, IL, USA

mridulsarkarbiz@gmail.com