Domain: Aerospace & Defense
Software Guidelines Assessment of Mission-Critical Systems Software for a Large Fortune 100 European Aerospace and Defense MNC
We performed a Complete Assessment of mission-critical software systems for coding-related anomalies, software design patterns, architectural patterns, and anti-patterns at various levels of granularities based on ISO-IEC 9126:25010 international standard. During this project, we designed and deployed this assessment documentation encompassing static analysis, dynamic analysis, software cloning, concurrency, performance, maintainability, reliability, scalability, portability and several such KPIs as a global standard across a large vertical of a European Aerospace and Defense Giant.
Domain: Space-Tech
Deployment of Edge ML Software on Satellite Processor for Earth Observation using AWS Ground Station and AWS Sagemaker, as part of a Global Hackathon sponsored by Deloitte.
This was a Global project with cross-sultural team members. We developed and designed Computer Vision and 3D Pattern / Object Detection Algorithms and Software using AWS Sagemaker services and MLOPs pipelines. We also optimized these algorithms and software on Satellite Edge Processor using AWS Ground Station service.
For this project, Use Cases under consideration were: Widfire/Bushfire detection at the Satellite Edge Node, Identification of illegal shipping vehicles at the Satellite Edge Node, Detection of Mangrove Depletion from UK coastlines.
Domain: Satellite Communications
Successful Development and Deployment of Data Acquisition and Image Analysis Software using 2 International Radio Telescope Systems [IRTS] of RAC and NCRA, TIFR, Govt. Of India.
In this project, we performed the First Interplanetary Scintillations (IPS) Observations which were conducted using the Giant Metrewave Radio Telescope (GMRT), largest radio telescope in the world at low frequencies. We developed and deployed software for data acquisition signal analysis (including noise removal using narrow-band and broad band RFI elimination techniques) using GMRT pulsar backend data acquisition system (DAC). We also performed image analysis of the data of 452 radio galaxies and quasars (QSOs) observed using the Ooty Radio Telescope (ORT) and estimated their 2-dimensional structure using 2-D Fourier Transform and Computer Vision (CLEAN Deconvolution) algorithms. We also developed and deployed a Data Visualization software for real-time monitoring of the wind profile and to perform Wind Tomography over the GMRT Array.
Domain: Chemical Industry
Data Mining, Data Visualization and Chemical Informatics Software Development and Support for Informatics and Robotics Global R&D Laboratories of Fortune 100 US MNC
We managed the Informatics and Robotics software of Symyx Corporation Suite of Products for Laboratory Automation of High Throughput Research Informatics (HTRi) and Robotics Laboratories at the Global Head Quarters of one of the largest Chemical Giant US MNC. We owned the development, modification and delivery of research computing software for HTRi and robotics, along with their supercomputing software systems and platforms in US, Europe and India. We developed statistical pattern recognition algorithms for computer vision patterns in images and videos of vials as part of the setup. We developed and deployed a Data Mining software for analysis of laboratory data. We worked on mathematical modeling for new materials development and integarted it with HTRi results through Chemo-Informatics and Chemometrics setups.
Domain: Mining & Construction
Predictive Maintenance of Assets on the Mining / Oil and Gas Field of Large Fortune 100 Mining / Oil and Gas Giant
We developed algorithms (such as Change Point Detection - CPD) and software for predictive maintenance of mining assets, and extensively used asset diagnostics, prognostics, and FMEA (failure-mode-effect analysis) techniques. We also integrated the predictions with Big Data Store on Cloud, and built data pipelines for real-time data acquisition using batch and streaming analytics modes. For this, we performed extensive analysis of time series data coming from sensors mounted on mining assets, and developed digital twins of the same using physics-based models, data-driven models and hybrid models for correct simulation of digital twins, before their deployment to the production system. We applied various algorithms in Machine and Deep Learning (using Deep Neural Networks) to refine the solution. The solution stood out in comparison with more then 15 similar cloud platforms and proved customer acceptance carried out via User Acceptance Testing (UAT).
Domain: IT/ITES
Engineering Analytics Software Platform for Product and Platform Engineering Vertical of a Multi-Billion Dollar Service based IT Giant
We developed a Generic and Customizable Software Platform using Cloud, Big Data and AI technologies to solve the Information Technology - Operations Technology (IT-OT) Integration Problem in the IT Service Industry. We developed reusable component based software platform capable of reusing AI/ML, Big Data Engineering, and Domain-Specific Software components for Energy and Utilities, Medical and Healthcare, Automotive, Oil and Gas and Mining Use Cases.
Domain: Industrial Automation/Smart Manufacturing
Proprietary Mesh Computing Algorithm and Hybrid Edge-Fog-Cloud IoT Architectural Solutions for One of the Largest Fortune 100 Industrial Automation US Giant
We developed Mesh Computing proprietary algorithm(s) for implementing distributed consensus and software redundancy in .a cluster of edge devices used in control systems for industrial automation. For this, we used OPC/UA and FDT standards to standardize the edge computing and ML algorithms / software with Fog and Cloud software and systems. We also implemented Centralized, Distributed and Decentralized Architectural Paradigms using Function-As-A-Service, phsPaxos, Blockchain and similar software frameworks and algorithms. We implemented these solutions from concept to delivery and deployment into production systems used for Industrial Automation.
Domain: BFSI
AI and Big Data Software Solutions for Commercial Cards, Commercial and Consumer Banking Divisions of one of the largest Fortune 100 US Banking Giant
We developed and deployed multiple Big Data Engineering and AI based software solutions for commercial cards, consumer/retail banking and commercial banking divisions of a US MNC Bank. For this engagement, the key Use Cases were: 1. Data Integration of multiple software assets of commercial banking by creating Single-Source-Of-Truth (SSOT) Data Lake pipelines using Spark, Scala, Java and Python software, 2. Automated Customer Onboarding using Generative AI models and software for document processing, optical character recognition, signature identification, font identification, 3. Multi-Terabyte Data Migration of Commercial Cards Siloed Data from legacy to RDBMS systems and from on-premise RDBMS to Cloud using batch and streaming analytics.
Domain: Healthcare & Pharma
Building One of the Largest Single Source of Truth [SSOT] as Multiple Data Architectural Paradigms for a European Pharma Giant
We performed Data Architectural Assessment of paradigms like Data Lake, Data Lakehouse, Delta Lake, Data Fabric and Data Mesh for organization-wide data assets and software of a large Global Pharma MNC. As part of the solutioning journey, we provided and implemented the right solution as a Hybrid Data Architectural Paradigm using Data Fabric, Data Lakehouse, Delta Lake and Data Mesh, as appropriate for multiple Business Verticals.
Domain: IT Consulting
Data Fabric Deployment for a BIG4 Consulting Firm using Microsoft Azure Cloud
We developed Multiple Big Data Software Frameworks for Large-scale Data pipelines across huge data assets and legacy software across a Big 4 Consulting Giant, sponsored as a long-term Co-R&D project by an Energy MNC. This included development, design and architecture and deployment of software frameworks for: Data Migration, Data Loading, Data Extraction, Data Transformation, Data Pre-processing, Data Storage, Data Profiling, Data Reconciliation, Metadata Management, Metadata-Driven Architecture, Data Trust Architecture, Data Security, Data Processing, Data Dictionary and Data Visualization.
Domain: BFSI
Full Stack Data & Application Software for Large Scale Data Processing for US Payment Solutions Fintech Giant
We developed, designed and deployed Full Stack Application Software and Data Pipelines / Solutions for large-scale real-time Big Data Processing and Management for a Payments Data Provider Giant MNC. For this, we used Distributed Message Broker Frameworks like Kafka along with Containerized deployments using Docker and Kubernetes services on Cloud Platforms. This scaled the application-level software for Asia-Pacific (APAC) region compatibility, considering the huge data in real-time passing through the payment cards ecosystem of this US payments and FinTech MNC.
Domain: Oil & Gas
Software Architecture and Design for Large Scale Data Processing Software Products and Systems for Upstream Technology of One of the Largest Fortune100 US Oil & Gas Giant
We provided software architecture, design and coding guidelines solutions for Big Data Processing systems used in Upstream ecosystem of multiple software products of a US Oil and Gas MNC. We consulted their US and Asia Divisions by assessing existing issues in their software products handling huge data in real-time. Improved the maintainability of huge legacy software systems and products by identifying object-oriented anti-patterns (God Class, High Coupling, Tradition Breaker, etc.), performance and concurrency anti-patterns; and rectified them without breaking the development and delivery lifecycle of the client, by passing all unit and integration tests.
Domain: Government Research Project
Niche Quantum-AI Algorithms for Analysis of Cosmology Datasets of NASA and ESA for Quantum Computing Applications Laboratory [QCAL], MIETY, Govt. Of India
We developed Hybrid Quantum-Classical [AI] Algorithms for detection of 3D shape (geometry, topology and algebraic properties) of large-scale structures of galaxies in the cosmic web of galaxies: data obtained from NASA/ESA datasets like Sloan Digital Sky Survey (SDSS) and Dark Energy Spectroscopic Survey (DESS) catalogues. We have implemented Quantum AI Algorithms on photonic, superconducting and trapped ion Quantum Hardware systems made available through AWS Bracket and Xanadu Cloud. As an extension to this project, we identified algorithms for Photogrammetry which can be applied to point clouds, 2D/3D mesh and voxel representations of telescopic / satellite data obtained from RADAR, LiDAR, SONAR, Optical, Infrared, and Hyperspectral imaging devices mounted on satellites and spacecrafts. We further utilized the analysis for Cosmic Cartography and 3D Simulation of the Cosmic Web, and also identified Quantum-AI Algorithms to identify anisotropy and inhomogeneity in the galaxy distribution in our universe using spherical harmonics and other similar methods. During the course of our work, we discovered Deep Neural Network Architecture alongside this work for Algebraic, Geometric and Topological Deep Learning applied on cosmology datasets, and more...
Domain: Healthcare & Pharma
Intelligent Chatbot Development and Integration for EU Region of a Global Pharma Giant
We developed, designed and deployed an Intelligent Chatbot using voice-to-text GCP API, Dialogflow framework, custom phonetic pattern recognition algorithms integrated with knowledge graph repositories / database at the backend, by implementing iterative refinement of the solution using multiple feedback loops. For this project, we utilized technologies like Conversational AI, Graph Databases and Generative AI to refine the solution and deployed it on light weight front-end on androis and iOS mobile operating systems. This was used in real-time for the Europe Market for the Pharma MNC clientele.
Domain: IT Product Engineering
Development of Code Analytics Software Product for an Indo-German Product Development IT Company
We developed, designed, architected and deployed multiple features of a Code Analytics Platform, capable of automatically identifying static and dynamic analysis issues in software systems. This product was compatible for analyzing software code in 12+ different programming languages. Integrated Big Data, Machine Learning and AI features into this software platform by developing toolchains for leading international research projects like: 1. DEviSE (Automatic Identification, Extraction and Reuse of Software Components from Software Repositories), and 2. QualiCORE (Automatic Identification of Concurrency Anti-Patterns in Software Systems), sponsored by the German Government. For building features into this product, we lead international research projects and collaborated with international universities / institutes like Karlsruhe Institute of Technology (KIT), Fraunhofer Research Institute, Petafuel, FZI, and Saarland University for the same.