|
The Infrastructure Management Services group at IRL-Delhi is involved in projects related to all aspects of systems management and service delivery as it relates to IT infrastructure. Our focus is on enterprise scale systems monitoring, resource optimization, capacity planning, root cause analysis, autonomic systems and adaptive systems and on methods and processes that allow these capabilities to be remotely delivered using a shared infrastructure. Examples of problems that we are addressing include:
- Data center monitoring at the component, service and business levels
- Data center power optimization
- Workload profiling
- Actuation (autonomously recovering from faults)
- Business Integrity and Compliance
- Root cause analysis in various contexts
Skills: Knowledge of one or more of distributed computing, systems management, autonomic systems, data mining, modeling.
Level: PhD/M-tech/MS/B-tech
|
|
The Information Management (IM) group at IRL is actively engaged in developing next-generation technologies in various areas such as data/text/speech analytics, context-oriented information integration, advanced business intelligence and human-computer interaction. The group has been building intelligent products and solutions by applying these technologies to address business problems in various industrial sectors, such as, financial, telecommunication and health-care.
In the area of context-oriented information integration, we have been developing novel techniques for loosely-coupled structured and unstructured data through symbiotic and semantically-disambiguated information in an enterprise. We are currently focusing on extending this technology to automatically discover different types of "entities" and relationships between them from unstructured data. The challenges involve learning context of different structured entities and relationships from imprecise and incomplete references in unstructured documents, fuzzy matching and search over structured data, and discovering the entity view automatically.
In the area of Information Extraction (IE) from unstructured data, our focus is on technologies which involve the identification of named entities and relationships among them. We are addressing several problems in this area that are crucial for a large-scale adoption of information extraction in practice, such as automatic discovery and engineering of effective features, use of disparate knowledge sources, optimal organization of IE rules, and scalable and efficient processes.
New paradigms of information extraction using visual layout of pages
Email analytics to automatically extract concepts from emails and classifying them into various categories using ML and rule based approaches.
RDBMS and high volume of data: exploring possible technology barriers of current RDBMS when managing terabytes of data.
Skills/Level: Ph.D. in Databases, Information Retrieval, Data Mining or related areas.
|
|
The Services Information and Analytics (SIA) department at IRL is involved in projects related to application of natural language processing, information retrieval, text analytics, data mining, machine translation, and speech technologies to service delivery problems. We deal with unstructured, noisy, and multi-lingual information sources such as Web pages, documents, e-mail, ticketing systems, team rooms, and IT monitoring systems to generate business intelligence and insights that can improve the quality, productivity and efficiency of IT and contact center services. We also build analytical models and solutions for knowledge process outsourcing, and tools for workforce optimization in IT services.
One of our focus areas is Services Analytics (SA). We are working on various applications of analytics for increasing efficiency as well as improving productivity and resource utilization in Services businesses. Some of the projects we are working on include improving the efficiency of resource utilization by automatically suggesting matches of people to projects, automatically filtering resumes of people applying for jobs, improved authentication through biometric fusion, etc.
In the area of contact center research, we are developing an array of technologies ranging from automatic evaluation of the linguistic abilities of applicants to customer satisfaction analysis involving heterogeneous contact center data sources (such as, audio conversations, agent logs and feedback forms). The goal is to transform contact center services from a labor-based model to an asset-based model and help business analysts generate business insights to improve the quality of the delivery and enable them to deliver high-value services. In the area of speech technologies, we are working on solutions that help people access information in various forms by overcoming barriers such as physical disabilities, lack of IT skills, and illiteracy,.. Examples of problems that we are addressing include information extraction from World-Wide Web using voice interactions and CSAT analysis to extract dissatisfaction drivers among customers.
Skills: Knowledge of machine learning, text mining, natural language processing, speech technologies, pattern recognition, data mining, information extraction/retrieval; RDBMS, Java/C++ for prototyping.
Level: PhD/M-tech/MS
|
|
The High Performance Computing (HPC) group at IRL is engaged in the design and analysis of cutting edge parallel programs and in improving the performance of engineering, scientific, and business applications on high performance platforms such as the IBM Blue Gene Supercomputer and the Cell processor. The group is focused on areas based on performance on multi-core processors, performance on large-scale supercomputers, performance on clusters, grid performance, medical imaging applications and parallel scalable algorithms for supercomputers. Examples of problems that we are addressing include:
- Optimization of benchmarks such as FFT and Transpose for novel supercoumputing (petaflop) architectures
- Parallel Algorithms for Medical Imaging applications such as 3D-CT reconstruction
- Parallel and distributed algorithms for text indexing and searching
- Parallel Algorithms for solving large, possibly ill-conditioned sparse matrices
Skills: Good knowledge of C/C++ and data structures; Principles of computer architecture,
Parallel and distributed processing desirable
Level: PhD/M-tech/MS/B-tech
|
|
Ongoing research effort in Programming Technologies and Software Engineering department includes productivity improvement tools for next-generation programming languages, and model-driven technologies for high-quality, industrial-strength service-oriented solution construction, evolution and delivery. These themes are inspired by following two trends:
- since scaling up the clock frequency of future processors is no longer a viable option, multi-core based systems will exploit parallelism more and more for increased performance.
- Componentization and standardization will further accelerate construction and evolution of business software in a distributed fashion, often by globally distributed teams enabled by new communication and collaboration technologies. We are looking for interns interested in applying program analysis, text analytics and software engineering principles for advancing our research agenda.
Skills: Knowledge of one or more of Java, Program Analysis, Software Engineering and Text Analytics.
Level: PhD/M-tech/MS/B-tech
|
|
The Telecom group at the IBM India Research Lab in Delhi and Bangalore executes innovative projects focused on the telecom industry. Some of the main focus areas are:
- Enabling IT for Emerging Economies utilizing the Mobile Platform
- Telecom Analytics and Business Intelligence
- Telecom Middleware
- Application Development on Mobile Devices
- Mobile Value Added Services including Mobile Commerce, Mobile Social Networking, Mobile Advertising, Mobile Payments etc.
- WiMAX/LTE (4G Wireless) and Multi-core parallel programming
- Network and Systems Management for Data Centers and Telecom Networks
Skills: Knowledge of one or more of Mobile Devices, Distributed Systems, Networking, Data Mining; Java/C++ for prototyping.
Level: PhD/M-tech/MS/B-tech
|
|
Analytics and Optimization covers activities in business optimization, machine learning, predictive modeling, and related business intelligence infrastructure and solutions. We are highly interdisciplinary in our approach, incorporating techniques from optimization, theory and algorithms, artificial intelligence, database management, stochastic analysis, pattern recognition and statistics. Examples of problems that we are addressing include:
- Emergency Management: Strategic and tactical resource deployment planning for managing Emergency events.
- Workforce Management: Resource deployment managers in a global service delivery organization must deploy hundreds of highly-skilled resources every day to new engagements. These deployment managers need help in identifying the best candidates and best use of GBSs resources in meeting the demands both tactically and strategically in a timely manner.
- Open Collaborative Research Project (with University collaborators): Goal in this project is to develop techniques analyze service interaction networks to improve overall service delivery process.
Skills: Trained in Optimization or Machine Learning. Programming skills in C/C++/Java is needed. Moreover, knowledge of tools/environments like MATLAB, Arena, AMPL/CPLEX is desirable, Linear & Nonlinear Optimization, Dynamic programming, exposure in Revenue Management
Level: PhD/M-Tech/MS
|
|
 |
|
