Harsh Multani Email & Phone Number

Berlin, Germany

Building The District VR : The portal for digital music entertainmentVR / Backend / Blockchain• designed the architecture of the product. It comprises of services from AWS.• worked with docker, docker-compose, & also setup CI/CD for the project.• wrote APIs using TypeScript which interacts with Postgres as a Database.• working on the interaction between backend server and Lukso Blockchain.• working on writing smart contracts using Solidity.

Munich, Bavaria, Germany

• customized backend code of Hyperledger BPA to cater to business needs.• worked with docker, docker-compose, & also setup CI/CD for the project.• developed UI pages for e-commerce applications using Angular.

Hyderabad, Telangana, India

NFT Marketplace : • developed a NFT Marketplace for minting, selling and auctioning of tokens. The creator of an art piece can mint it’s token, sell it and auction it. Also, when it is sold to another user and when this user sells it forward, royalties would be paid to the author. ERC-721 and ERC-1155 were the token standards used.• created a private ethereum network using Geth. Also, created a private IPFS network to store the art files in a distributed way.• wrote Solidity smart contracts and interacted them with private ethereum network, ethereum testnets and polygon testnet.• customized functionality of minting, selling, buying of tokens to cater to business needs. Also, coded functionality of royalty payments.

Pune, Maharashtra, India

DT:• created a decentralized Identity Private Permissoned network of the future using Hyperledger Fabric and Aries. • An application that enables the individual to verify his/her credentials without even revealing it to the verifier through zero knowledge proofs. • architecture comprises of Hyperledger Fabric, Aries Agents, Kafka, microservices through docker container, Redis, Postgres, MariaDB.• worked on writing DAML Smart Contracts, customizing Aries Framework dotnet, writing Java services, Kafka and docker.• did research on ZKP-CL, BBS+ cryptography methods.• increased performance by tuning Kafka and adding postgres as a cache to Fabric.Open Source Contribution :Endorser Protocol :• contributed Endorser Protocol to Hyperledger Aries Cloud agent Python.• The protocol was developed so that not everyone is allowed to create transactions in the Hyperledger Indy ledger and it should only be created after it has been cryptographically signed by a particular endorser.• coded Endorser Protocol flow in Aries Cloud Agent Python.• worked on Hyperledger Indy, Aries, Ursa and Python.

San Francisco, California, United States

Open Source Contribution :Aries - Fabric Wrapper:• worked as a Mentee under Linux Foundation and solved few interoperability problems which would then enable Aries Framework Javascript to support Indy and Fabric as a Blockchain Ledger.• wrote chaincode and client application in Typescript.• worked on integrating client application with AFJ.

Hyderabad Area, India

Conversational AI :• setup a conversational AI system that responds to the input query.• Architecture comprises of Hadoop Stack, HDFS, Spark and Kafka, along with some NLP models.• wrote spark jobs in Python / Scala to do the data pre-processing.• built NLP models for text classification, with the use of BERT embeddings.• NLP models for Text Classification, Named Entity Recognition, Question – Answering systems were built• created an in-app chat feature that allows the user to have a conversation with a live agent.• REST APIs were built using Flask, FastAPI, Django. gRPC was also used.OCR Project :• demonstrated how OCR technology can be used for complex tabular data extraction from scanned or readable PDF’s.• The PDFs had tables between columns, vertical table and also there were multilingual texts in some tables.• Use of Google Tesseract, Amazon Textract, Azure Form Recognizer & Azure Computer Vision.• Custom ML models were also built using Azure Form Recognizer.• Image Analysis was done using OpenCV, ImageAI.Smart Tracking using Blockchain (POC) :• Supply chain tracking of parts of aircraft - tracking when the parts are manufactured by the contractor - upto the quality certification and then finally – when the shipments are delivered.• All the transactions are being input into a Hyperledger Fabric Blockchain.• Node.js was used to develop the POC and REST APIs were developed using Loopback.• Bosch XDK 110 sensor was used, and the data from the sensor was input into the blockchain using Eclipse MITA.Heat Exchanger Quality prediction (POC) :• predictive analytics to measure the efficiency of Heat Exchanger.• The IOT data that had to be taken from sensors, was taken using simulation, as there was covid-19 and we had to WFH.• different simulators like Bevywise, NodeRed, CupCarbon, Kepware have been tested.• The data from PLC was then processed into the on-premise database.• SVM, ANN, Random Forest models were built using this data.

Bachelor'S Degree, Computer Science

Master Of Science - Ms, Data Science