Looking for a job

Ritesh Panjwani

@riteshpanjwani

Message

Software Development Engineer III at 3SC, building AI agents and approval workflows that save 1,000 hours monthly.

India

Message

At 3SC, I built the SCAI Agent Platform, where clients use custom agents grounded in more than 100 platform APIs. It’s live with over 10 clients and saves about 1,000 hours a month.

I also built an end-to-end approval workflow microservice that processes millions of approvals, and helped cut sign-off times from days to hours. Earlier, I developed Java backends at Locus and data science, NLP, and machine learning systems at Loylty Rewardz, Rakuten, and the Center For Indian Language Technology, IIT Bombay.

Experience

Work history, roles, and key accomplishments

Current

Software Development Engineer III

Current

3SC (SS Supply Chain Solutions)

Oct 2023 - Present (3 years)

Built the SCAI Agent Platform (agentic AI) with LangGraph and pgvector RAG, serving 10+ clients and 50+ users. Developed the Approval Workflow microservice end-to-end, processing millions of approvals across 9 plan, norm, and order types in 5 production deployments.

LangGraph Pgvector RAG Django Rest Celery PostgreSQL Redis Python SQL AI Agentic AI Microservices CI CD Pytest

Backend Developer

Locus

Nov 2021 - Jul 2023 (1 year 8 months)

Built robust backend systems in Java and Spring framework, optimizing order upload for a 5x reduction in upload time. Conducted code reviews and led on-call incident response.

Java Spring Framework SQL Code Review Incident Management

Data Scientist I

Loylty Rewardz

Jun 2019 - Oct 2021 (2 years 4 months)

Developed algorithms to extract information from over 8 billion unclean bank transaction narrations using NLP, achieving over 23% improvement. Built a Python Flask tool to automate marketing campaigns, cutting processing time 10x and saving ~84 man-hours.

Python Flask NLP Kafka MariaDB Greenplum XGBoost Machine Learning Data Pipelines

Technical Intern

Rakuten

Oct 2018 - Mar 2019 (5 months)

Developed a pipeline in Spark to ingest, preprocess, and create machine learning features from one billion credit card transactions. Classified a huge unlabeled e-commerce product catalogue in semi-automated fashion using pre-trained image classifiers.

Spark Machine Learning Data Processing Image Classification

Research Engineer

Center For Indian Language Technology, IIT Bombay

Nov 2016 - Sep 2018 (1 year 10 months)

Developed a Python-based library to access synsets, glosses, and lexico-semantic relations for 18 Indian language WordNets, published in GWC2018. Compared Convolutional vs. Recurrent Neural Networks for machine translation, published in WAT2017.

Python NLP Neural Networks LSTM Machine Translation WordNet