Large organizations respond to numerous volumes of Request for Proposals (RFPs) annually. The process of responding to RFPs is largely a manual and time-, effort-, and intellect-intensive endeavor, and vulnerable to errors. To bridge these gaps, we present RFPAnaFit, a tool designed for RFP Analysis, Fitment evaluation, and automated response generation. RFPAnaFit provides an automated solution for generating responses to RFPs. RFP typically comprises a set of questions that outline the requirements for various capabilities that need to be addressed. It leverages product capability documents and uses multiple techniques to assess the fitment of RFP requirements against product capabilities and generate RFP response document. We have tested RFPAnaFit on a real-world, large-sized product with two RFPs, demonstrating its effectiveness and providing valuable insights. While the findings are shared within the context of an industry-specific product, we believe that researchers and practitioners will find the lessons learned from this approach applicable to other domains as well.

Requirement engineering in many IT services industries continues to be document-centric and heavily manual. Requirements specification documents contain details of product features, process flows, activities, rules, parameters, etc. Intricate knowledge of dependencies between these specification elements is necessary for effective change impact analysis. In document-centric change impact analysis, Subject Matter Experts (SMEs) have to search for information across multiple requirements specification documents. Especially with business products having multiple customized product implementation scenarios, it is crucial to consider the context (client/ operating market/ geography) when analyzing the change request (CR), that makes the overall CR impact analysis process a time-, effort- and intellect-intensive endeavor, and vulnerable to errors. To overcome these challenges, we propose Natural Language Processing (NLP) and Model-Driven Engineering (MDE) based, automated Contextual Change Request Impact Analysis (ContCRIA). ContCRIA automates change impact analysis and generates a detailed CR impact analysis report. In this paper, we describe the overall approach of ContCRIA and its application on two real-world products thus bringing out its efficacy as well as lessons learned. Though the findings are shared in the specific context of two industry products, we believe researchers, practitioners, and tool vendors will find the takeaways from this approach and experience applicable in other contexts too.

Enterprise application development faces significant challenges, with each phase of the software development life cycle (SDLC) requiring experts with specific skills. The expertise of the individuals involved, greatly affects the quality and speed of work in each phase. The large size and complexity of modern software systems further exacerbates these problems. Recently, there has been a growing interest in using Generative AI (GenAI) techniques for software engineering tasks. GenAI can help Subject Matter Experts (SMEs) work more efficiently and can help in overcoming skill barriers. By leveraging GenAI, SMEs can save significant time and effort. This paper introduces meta-model based prompting approach to generate enterprise application code leveraging large language models (LLMs). Prompts help in the refinement of input requirements into refined requirements and design specifications using LLMs, ultimately generating code from these specifications. We share our approach and results of applying approach to generate small yet complex applications.

Medical Writing is an art of writing scientific documents which includes regulatory and research-related content. To obtain approval for marketing new medicines, pharmaceutical companies are obligated to provide drug authorities with a huge volume of documents related to clinical trials. Creating these clinical trial documents is a time, effort, and skill-intensive process as the required information exists in fragmented form distributed across various information sources. To overcome these challenges in medical writing, we propose Automated Medical Writing tool (AutoMW). AutoMW enables the digitalization of information from different sources of information using a meta-model-based approach and leverages these models for the automated generation of clinical trial documents as per the regulatory authority document templates. This paper describes the approach and illustrates its utility and efficacy in real-world clinical trial application of two use cases - breast cancer, and diabetes.

Typically, business applications have complex and extensive functionality. Often, rules are scattered throughout the application documentation and code, making it challenging to modify them. Because business rules change more frequently, it is preferable to externalize the rules and move them outside the application. The requirements specification can serve as a valuable source for extracting and classifying rules that can further aid in assigning the rules to the appropriate teams based on their areas of expertise and externalizing the rules in the implementation. We introduce RClassify, which extracts and classifies rules from requirements specifications spread across several NL documents into eight different business rule classes. RClassify combines NLP and ML-based classification approaches to improve classification accuracy. We discuss the implementation of this approach in three real-world large-size and complex products, demonstrating its effectiveness, experience, and lessons learned. While the findings are presented in the specific context of three industry products, we believe that researchers, practitioners, and tool vendors will find the insights and experiences gained from this approach applicable in other contexts.

Requirement engineering in many IT services industries continues to be a document-centric and heavily manual activity, relying on the expertise of business analysts. Requirement specification documents contain details of product features, process flows, activities, rules, parameters, etc. Intricate knowledge of dependencies between these specification elements is necessary for carrying out the effective evolution of the product over time. Today, Business Analysts (BA) are forced to recourse to keyword-based search across multiple requirement specification documents which is a time-, effort-and intellect-intensive endeavor, and vulnerable to the errors of omission and commission. To overcome these lacunae, we propose DizSpec, an automated approach for digitalizing the requirement specification documents into a model form through automatic extraction of specification model elements and the various dependencies between them. The proposed approach creates a digital thread providing machine-processable traceability from product features to its specification elements. It also provides an easy natural language querying mechanism to generate traceability and impact analysis reports of interest. In this paper, we describe the application of this approach to two real-world products thus bringing out its efficacy as well as lessons learned from this transformation journey of the document-centric process to a model-centric and automated process. Though the findings are shared in the specific context of two industry products, we believe, researchers, practitioners, and tool vendors will find the takeaways from this approach and experience applicable in other contexts too.

Software quality has become the lever of differentiation in today's competitive marketplace. Quality at speed is the customer demand and automation is the biggest bottleneck holding the evolution of quality function. Increased levels of automation and intelligence in software engineering are the emerging trends across the IT field. As systems and software processes guide the life cycle activities and are the vehicles for building quality, it is necessary to look at the process infrastructure for the extent of process automation support provided and the digital enablement. This paper maps out the existing process infrastructure support in industry practice and proposes a roadmap for digital re-imagination of software and systems processes. Harmonizing the quality engineering themes with digital technologies, we propose a framework for building an intelligent software process infrastructure, iSPIN that can help in digital re-imagination of software and systems lifecycle processes. The framework has been implemented using digital technologies and has been piloted with one of the industry business unit for re-imagination of "proposal process". The proposed iSPIN framework will help in unprecedented automation and quality engineering at each process step and paves the way towards realizing the dictums of "Quality at Speed" and "Digital transformation of Software Process".

With internet becoming increasingly pervasive, data analytics is playing increasingly critical role in the business. Data to be analyzed exists in large quantity and in multiple formats. Many technologies exist to support Big Data analytics. However, they remain somewhat of a challenge for average developer to use. It's been seen that model-driven development (MDD) approach can eliminate accidental complexity to a large extent. We discuss MDD approach for development of MapReduce based Big Data applications, its efficacy and lessons learnt.

We share experience in supporting development and evolution of a large banking product using a homegrown model driven engineering (MDE) toolset. We discuss improvements that needed to be introduced in the MDE toolset to support collaborative development with teams distributed across different geographical locations. Though experience is shared in a specific context, we believe, MDE researchers, enthusiasts, practitioners and tool vendors will find the takeaways from this experience applicable even in a more general context of large scale software development.