CarbonSustain, Inc.5 Year Plan: Decarbonize your operations with AI
PROJECT DESCRIPTION: 
CarbonSustain is carbon emissions accounting and AI driven insights as a service for SMBs. CarbonSustain streamlines scopes 1,2, & 3 emissions, boosts savings, and helps companies elevate their brand. CarbonSustain offers AI-driven actionable insights facilitating a cost-effective journey to Net Zero while enhancing the environmental aspect of its brand image. Powered by legislation tailwinds including the US Inflation Reduction Act and California's Climate Accountability Package, CarbonSustain delivers a service akin to a TurboTax for carbon emissions reporting for small & medium businesses.


We're building a Decarbonize your journey to Net Zero with AI. After doing the carbon emission accounting for a SMB, we will have the entire value chain mapped out. The AI module we're building requires a lot of data analysis and specific business rules specified. We want the discovery team to build in hundreds of business rules to help a business reduce costs and as a byproduct reduce carbon emissions. To create a classification model with AI to find the cluster of companies that are most comparable to that one. Once cluster is identified, use AI to determine how those companies were able to improve their sustainability score & reduce their carbon footprint. The journey should include a 5 year timeframe.


PROJECT TIMELINE: 
Each Sprint will be two weeks: Sprint 1: Onboarding, Data Discovery, Data Cleaning, Data Pipeline (Python, Excel). Sprint 2: Model training, experimentation with ML algorithms for classification and regression, develop Chatbot (Azure). Sprint 3: Model orchestration, develop API to support running the models and retrieving predictions from the various models by providing the attributes. Sprint 4: Integrate the APIs and models into the frontend and save appropriate attributes in the DB (Angular, AWS, RDS, MySQL, NodeJS). Sprint 5: Optimization sprint, add more datasources, reduce data loss (drops), increase accuracy of models. Enhanced frontend UI for predictive data. Sprint 6: Test Driven Development (TDD), focus on DevOps, write Unit testing, automated testing to ensure the veracity of the data.

PROJECT WORKFLOW: 
We will be meeting on UC Berkeley campus twice a week, Monday and Thursday on various co working spaces: CITRIS Climate Innovation space at Sutardja Dai Hall, Blum Hall, and also Haas. We have interns coming in from the MET program, and many from SkyDeck as well. The other days we will be remote and will have communication via slack, whatsapp, Zoom, text, and email. We will track bugs through Jira and use Git for checkin with pull requests.

DELIVERABLES: 
The functioning AI/ML models to do the various AI applications as required to develop the decarbonize your journey in AI 5 year plan plugged into the live production website.

PREFERRED APPLICANT SKILLSET:
  • Python - Beginner/Intermediate/Advanced
  • SQL - Beginner/Intermediate/Advanced
  • EDA - Beginner/Intermediate/Advanced
  • Data Visualization - Beginner/Intermediate/Advanced
  • NLP - Beginner/Intermediate/Advanced
  • Machine/Deep Learning - Beginner/Intermediate/Advanced
  • Cloud Computing - Beginner/Intermediate/Advanced
PREFERRED ADDITIONAL SKILLSET: 
AWS, RDS, Angular, NodeJS, RDS, Azure, Chatbot functions, Linear Regression, Random Forest, model orchestration, restful API, OpenAI, Pandas, Python, Excel, big data, MySQL, Web3: XRPL, Smart Contracts, Carbon Tokens, Carbon Emission Tokens (CET), Toucan Protocol, Polygon, Celo

Depending on the project, students can expect to spend anywhere between 6-15 hours per week. While backgrounds in STEM are all useful, we believe that diverse perspectives and experiences are the most essential qualities for successful research. 
Email *
Next
Clear form
Never submit passwords through Google Forms.
This form was created inside of UC Berkeley.

Does this form look suspicious? Report