VP of AI

About NanoNets

Automatic Data Extraction

Tech

### Some of the interesting things our backend team has shipped - Compile python code into C which could be imported into golang and then shipped as binary for on premise systems - Autoscale GPU dependent services with kubernetes with a custom metric - Displaying machine learning metrics in simplified ways to end users so they can act based on those metrics - Building large number and variety of integrations with relatively generic interface like salesforce, quickbooks, RPA's, external databases - Process large number of files in highly distributed manner in golang ### Some of the interesting things our frontend team has shipped - Ability for users to annotate documents so AI can learn which fields to extract - Displaying machine learning metrics in simplified ways to end users so they can act based on those metrics - Letting users build complex visual workflows around our API in our product. - Let users visualize complex ML metrics in a very simple and intuitive way Our stack: - Databases - Cassandra DB - Postgres/MySQL - Backend - Golang for API and other microservices - Python for Machine learning (Tensorflow, Pytorch) - Frontend - React, Typescript - Mobx - Cloud Providers - AWS - GCP for ML heavy workload - Monitoring/Alerting - ELK for logging - Prometheus for Monitoring - Graphana for dashboards - Orchestration - Kubernetes - DevOps - Jenkins for CI/CD

The role

Nanonets has a vision to automate the most complex, time-consuming processes inside companies which take away hours of valuation time from employees. Our workflow automation platform is powered with the latest models that can make sense of even unstructured data like documents. Our client footprint spans across brands such as Toyota, Boston Scientific, Bill.com and Entergy to name a few enabling businesses across a myriad of industries to unlock the potential of their visual and textual data. Some of the technical challenges we deal with are fine-tuning SOTA VLM’s for workflow automation, using generating architectures for their superior emergent behaviour to understand the world but also constraining those architecture to reduce hallucination and generate structured workflows which can be used in completely automated manner. We recently announced a series B round of $29 million in funding by Accel and are backed by the likes of existing investors including Elevation Capital & YCombinator. This infusion of capital underscores our commitment to driving innovation and expanding our reach in delivering cutting-edge AI solutions to businesses worldwide. Read about the release here: https://www.forbes.com/sites/davidprosser/2024/03/12/why-enterprises-are-learning-to-love-nanonets-automation/?sh=6d79ec8f3ca1 https://techcrunch.com/2024/03/12/nanonets-funding-accel-india/amp/ ### What We Expect From You * Strong Machine Learning & DL concepts. * Strong command in low-level operations involved in building architectures like Transformers, Efficientnet, ViT etc., and experience in implementing those in pytorch/jax/tensorflow. * Experience with the latest semi-supervised, unsupervised and few shot architectures in Deep Learning methods in NLP/CV domain * Strong command in probability and statistics. * Strong programming skills. * Have previously shipped something of significance, either implemented some paper or made significant changes in an existing architecture etc ### **Ideal candidate should have the following skillset** * Experience building and deploying systems * Experience with Theano/Torch/Caffe/Keras/Python/Tensorflow all useful * Experience writing production software would be a plus * The ideal candidate should have developed their own DL architectures apart from using open source architectures. * Ideal candidate would have extensive experience with computer vision applications. ### **Interesting Projects Other Senior DL Engineers Have Completed** * Deployed large scale multi-modal architectures that can understand both text and images really well. * Built an auto-ML platform that can automatically select best architecture, fine-tuning method based on type and amount of data. * Best in the world models to process documents like invoices, receipts, passports, driving licenses, etc * Hierarchical information extraction from documents. Robust modeling for the tree-like structure of sections inside sections in documents. * Extracting complex tables — wrapped around tables, multiple fields in a single column, cells spanning multiple columns, tables in warped images, etc. * Enabling few-shots learning by SOTA finetuning techniques. Candidate should have experience working on Deep Learning with an engineering degree from a top tier institute Thinking of applying? Try our [resume builder](https://resume.nanonets.com/)— it's free, fast, and tailored to help you stand out.

Skills

• TensorFlow
• Torch/PyTorch