Jyotsha Kumar

Software Engineer

Jersey City, New Jersey, United States3 yrs 10 mos experience

AI EnabledAI ML Practitioner

Key Highlights

Led teams to optimize HVAC energy use by 10%
Developed NLP models saving over $1.5 million annually
Expert in backend systems and ML infrastructure

Stackforce AI infers this person is a Backend Engineer specializing in IoT and Machine Learning solutions.

Contact

Skills

Core Skills

Backend DevelopmentMachine LearningMicroservices

Other Skills

AIAdobe Premiere ProCCassandraCreative Non-fictionCreative WritingFinal Cut ProFlaskHVACHaskellHeat TransferIoTJDBCJavaJavaScript

About

I build backend systems and ML infrastructure that run in production. I’ve led cross-functional teams to deliver IoT + ML control loops (ingesting 1,000+ devices and ~500GB/day) that cut HVAC energy use by up to 10%. My strengths are distributed systems, data pipelines, model serving and production reliability — I’m strongest on backend and systems work. Open to SDE-II roles where I can design scalable systems and mentor small teams.

Experience

3 yrs 10 mos

Total Experience

1 yr 3 mos

Average Tenure

Current Experience

Bert labs

Senior Research Engineer

Jan 2023 – Jul 2023 · 6 mos · Bengaluru, Karnataka, India · On-site

1. Co-Managed a team of backend developers & data scientists to build a solution using IoT-AI based control loop for optimising HVACs (thermal and ventilation control unit) in commercial (offices, hotels) & industrial buildings (refineries, pharmaceutical clean rooms) reducing carbon emissions by up to 10%
2. Designed and created pipeline using node.js/javascript for adding data real time sensor data to mongoDB database followed by cleaning and preprocessing with Python to insert in Cassandra time series database
3. Developed and deployed a continuous action space reinforcement learning model using soft actor critic algorithm in Python using PyTorch to achieve 8-10% cost savings and reduction in carbon emissions

IoTAIHVACNode.jsJavaScriptPython+6

Itribe

Software Engineer

Jan 2022 – Nov 2022 · 10 mos

1. Led a team of 5 people in developing backend microservices (Python, Node.js) & frontend dashboard (JavaScript, React.js) for large telegram channels to view, analyse & manage subscriptions
2. Gathered requirements by understanding client needs and use cases, brainstormed plan of action and minimum viable product to satisfy clients, delivering the first working version in 3 weeks
3. Developed secure payments channel for clients in backend with telegram bot for subscriber side interaction via telegram

PythonNode.jsJavaScriptReact.jsBackend DevelopmentMicroservices

Citi

Software Analyst

Jul 2019 – Jan 2022 · 2 yrs 6 mos · Pune, Maharashtra, India

1. Developed natural language processing model for classification of financial securities based on transaction receipts using PyTorch in Python, removing manual touch points, in turn reducing company expenses by over 1.5 million dollars in yearly equivalent work-hours
2. Scaled in house Java caching solution to increase databases covered by 20%, implemented least recently used algorithm for cache
3. Created RESTFUL apis in backend using Python(Flask) and Java(JDBC) for serving front end dashboards and machine learning models data from Oracle and mongoDB databases

Natural Language ProcessingPyTorchPythonJavaFlaskJDBC+3