CIP Projects Fall 2025

Sponsor Introduction: Established in 1893, Finow Rohrsysteme GmbH is a longstanding manufacturer of high-quality piping systems and components. The company serves various industries, including power generation, oil and gas, offshore wind energy, and construction. Their expertise encompasses inductive and cold bending, heat treatment, cutting, welding, and mechanical processing. Notably, Finow Rohrsysteme has contributed to significant projects such as the Nord Stream pipeline and the Expo 2020 in Dubai. In 2022, the company was acquired by Technoenergy AG 

Project Description:  Finow Rohrsysteme GmbH is undertaking a strategic initiative to explore the value of Artificial Intelligence (AI) in core areas of business operations. This initiative was launched by the Executive Council and is coordinated across departments with the goal of building internal capability, identifying practical applications, and improving operational efficiency through intelligent automation. 

The program is structured around three pillars: 

• Leadership – Development of a company-wide AI vision, responsible tool selection, data strategy, and governance. 

• Lab – Practical experimentation through pilot projects, including this collaboration with the Mack Institute. 

• Crowd – Broad employee involvement to identify bottom-up use cases and promote innovation. 

The initiative is actively supported by the company’s AI Officer, digital transformation leaders, and engineering department. In first step we would like to discuss the strategy as well and maybe to make the first study with Mack Institute describing the strategy. In parallel we can work on the specific Lab direction “Pilot 3.3 Assistant for Designers” 

Business Context 

As a manufacturer of high-specification industrial piping systems, Finow’s engineering department is central to operations. Engineers are responsible for interpreting customer requirements, creating technical documentation, preparing quotations, and ensuring compliance with norms and standards. These activities are knowledge-intensive but also highly repetitive and time-consuming. The potential for intelligent support systems is significant. 

Objective 

To design, validate, and prototype a digital AI assistant that supports design engineers by automating and structuring key documentation and communication tasks in early-phase project and offer preparation. 

• In which engineering workflows can AI reduce repetitive or manual effort? 
• What document types and templates are most suitable for AI-supported generation? 
• How can previous project data or quotation documents be used as training or reference inputs? 
• What is the right level of AI intervention without compromising quality and control? 
• How should the solution be embedded in Finow’s Microsoft 365 environment? 

Sponsor Introduction:
LLR Partners is a lower middle market private equity firm that partners with growth-stage companies to build value through sector expertise, operational excellence, and long-term collaboration. In pursuit of deeper, faster, and more scalable insights, LLR is investing in foundational data infrastructure to support its diligence and underwriting processes—transforming how data is sourced, processed, and activated. This initiative lays the groundwork for eventual agent development and agent-based orchestration and AI-assisted decision-making, positioning LLR to stay ahead of market evolution. 

Why:
In private equity, time is critical. The majority of diligence data LLR receives—financial models, CIMs, management decks, contracts, and third-party reports—arrives in unstructured formats (Excel, Word, PowerPoint, PDF). Currently, this content is stored in Egnyte, then manually translated into other unstructured modalities like (Excel, Word, PowerPoint, PDF) and formats for validation, review, and decision-making. 

To scale this process and prepare for a future where intelligent agents assist in diligence and underwriting, LLR needs a robust data layer that can ingest, validate, and structure unstructured content at scale. By leveraging its existing platform architecture and aligning with its vision for agents and orchestrated agent swarms, this project will help unlock insights faster, reduce friction, and fundamentally reimagine the investment process. 

What:
This MBA project will focus on defining and designing the data layer that bridges unstructured data ingestion with downstream consumption across operational models, data science, reporting, and self-service. It will build upon LLR’s existing architecture (as represented in the provided platform diagram) to enable: 

• Ingestion from Egnyte & Other Sources: Define how files move into the “data lake from the original source data” layer. What is the optimal raw to staging to processed architecture and workflow? In many instances there will be multiple versions of files and redundant data in the original data source. 

• Refinery & Structuring Layer: Identify tools and methods for parsing and extracting meaning from PDFs, Excel models, Word documents, and PPT decks including gpts, open source libraries preferably in python, and paid solutions. 

• Event, Batch & Designed Data Processing: Map how these inputs flow into batch jobs, real-time streams (e.g., flagged events), and curated datasets usable by stakeholders. 

• Registry & Governance Frameworks: Ensure versioning, data lineage, and access control to build auditability and trust in structured outputs. 

• Subscriber Enablement: Define how structured data powers applications, dashboards, and future AI agents used across the firm. What is the ideal workflow for end users? 

How: 

Workflow Mapping & Problem Framing 

• Interview internal teams to document the full lifecycle of diligence data: from receipt to collaboration and analysis. 

• Identify bottlenecks in translation, validation, and reuse of data from unstructured formats. 

Tool & Technology Evaluation 

• Research leading document intelligence tools (e.g., Azure AI Document Intelligence, Google Document AI, Instabase). 

• Evaluate orchestration layers that align with your platform design (e.g., Apache Airflow, Prefect, Dagster, Inngest). 

• Assess metadata tagging and governance platforms (e.g., Alation, Collibra) for integration with registry components. 

Target Architecture Design 

• Propose a refined architecture for the “refineries” and “designed data” components. 

• Define how data flows from Egnyte through the data lake and is shaped for subscribers (applications, agents, dashboards). 

AI Agent Readiness Planning 

• Identify two to three use cases (e.g., VDR Data Pack translation to LLR Data Pack, document summarization, Investment Modeling). 

• Define data requirements and output formats for enabling these use cases with future AI agents. 

Governance & Scaling Plan 

• Recommend protocols for data stewardship, access management, and auditability. 

• Suggest a roadmap for scaling from a pilot (e.g., processing Excel models) to a full-scale system. 

Who:
This project will be conducted by MBA students in collaboration with LLR’s investment and technology teams. Ideal candidates will have experience or coursework in digital transformation, data architecture, private equity, or AI strategy. Familiarity with unstructured data handling, data pipelines, or AI engineering is a plus. 

Deliverables: 

• Presentation Deck: A concise yet comprehensive strategic design of the data layer, showing how it supports unstructured data transformation and AI-readiness. 

• Narrative Report: A 5–7 page document outlining key findings, technology evaluation, design rationale, and roadmap. 

• Optional Materials: 

• • Vendor shortlist and comparison for document intelligence/refineries 

• • Sample data flow mapping from Egnyte → Data Lake → Designed Data 

• • Mockup of a diligence agent using this infrastructure 

Success Metrics: 

• Efficiency Gains: Reduction in time to process and validate diligence materials. 
• Data Reusability: Increased ability to reuse structured data across different investment scenarios 
• Agent Enablement: Clear line of sight to future AI agent use cases with foundational data requirements met. 
• Strategic Alignment: Alignment with LLR’s platform architecture and long-term investment enablement goals. 

Sponsor Introduction: Willis Towers Watson (WTW) is a global advisory, broking, and solutions company that helps clients around the world turn risk into a path for growth. Headquartered in London, WTW operates in over 140 countries and employs tens of thousands of professionals. 

The firm was formed in 2016 through the merger of Willis Group and Towers Watson. WTW provides services across four key business segments: Human Capital and Benefits, Corporate Risk and Broking, Investment, Risk and Reinsurance, and Benefits Delivery and Administration. Their work spans actuarial consulting, employee benefits, risk management, and insurance brokerage. WTW is widely recognized for its data-driven insights and analytical tools that help organizations improve performance and manage uncertainty. b

Project Description:  

Why: The digital economy and digitization have grown exponentially with the proliferation and explosion of new technologies, means of communication, and ways of doing business.  With the advent of the Internet in the 90s, companies underwent a fundamental shift in how they dealt with their customers and clients.  Now, with societies now entering the artificial intelligence age, specifically GenAI, companies are again having to reassess how they approach consultancy services to deliver maximum impact for their clients.  This project stems from wanting to understand how digital disruption will affect the value proposition of existing insurance and consultancy services, the way companies interact with clients and prospects, and how to pinpoint potential client targets.     

 What: This project will delve into various aspects of digital disruption and what that means for companies who wish to have a leg up on the competition. Key focus areas include: 

• Looking at the existing insurance and consultancy services landscape – and what the future of that landscape may look like. 

• Examining the impact of AI on consultancies across a wide range of fields and disciplines. 

• Determining how GenAI may affect consultancy services in the coming 2-5 years. 

• Pinpointing how leading AI companies such as Microsoft, Nvidia, and Alphabet are incorporating products and services into their clients’ business offerings. 

How (Methods): 

• Data Analysis: Utilizing data on digital disruption across specific industries over time.

• Comparative, Historical Studies: Case study analysis examining different industries and their approaches to digital disruption.

• Interviews and Surveys: Engaging with experts in AI, risk management, the consultancy economy, and business models.

• Literature Review: Analyzing existing research on digital disruption.

Who (Qualification): 

Ideal candidates for this project are professionals with a blend of expertise in risk management, cyber security, technology, and business studies.

Sponsor Introduction: Willis Towers Watson (WTW) is a global advisory, broking, and solutions company that helps clients around the world turn risk into a path for growth. Headquartered in London, WTW operates in over 140 countries and employs tens of thousands of professionals. 

The firm was formed in 2016 through the merger of Willis Group and Towers Watson. WTW provides services across four key business segments: Human Capital and Benefits, Corporate Risk and Broking, Investment, Risk and Reinsurance, and Benefits Delivery and Administration. Their work spans actuarial consulting, employee benefits, risk management, and insurance brokerage. WTW is widely recognized for its data-driven insights and analytical tools that help organizations improve performance and manage uncertainty. 

Project Description:  

Why: As enterprises come to grips with the age of artificial intelligence (AI), they are now having to deal with the risks that AI poses to their operations, work culture, and client offerings. More than half of Fortune 500 companies now cite AI as a business risk, with annual reports published with the SEC witnessing a nearly 500% increase in highlighting AI risk since 2022.  The rise of chatbots and LLMs is prompting organizations to reexamine longstanding business models, operational strategies, and client engagement frameworks in an effort to integrate and embed AI adoption across the entire enterprise.  Businesses are now confronted with the harsh reality of being able to adopt AI in a manner that will be consistent with company culture, market advantages, and regulatory outlines while simultaneously providing them with a competitive edge.  This project is rooted in exploring how AI adoption rates are unfolding, what makes AI adoption successful, what industries are implementing AI adoption the best, what technologies are being used, and what are the largest, most significant hindrances in quicker and more in-depth AI adoption.   

What: This project will delve into various aspects of AI adoption and the technologies/skills that make such adoption successful (or just as importantly, unsuccessful).  With BCG estimating that only 4% of companies are developing cutting edge AI capabilities across functions and generating value, this endeavor will hopefully focus on the following key areas of research:   

• Looking at what forms AI adoption takes, and the metrics and benchmarks used to determine success and failure. 

• Examining AI adoption for cross-sectoral/industry lines, with a particular emphasis on large-scale enterprise adoption for both large organizations and SMEs. 

• Determining which AI technologies are the most conducive to successful AI adoption. 

• Pinpointing how leading technology companies such as Microsoft, Nvidia, Amazon, and others are incorporating AI adoption across their organizations, and what lessons, if any, this could provide for other companies and industries. 

How (Methods): 

• Data Analysis: Utilizing data on AI adoption rates across specific industries over the last 2-4 years. 

• KPIs and Models:  Determining whether models such as the AI Adoption and Maturity Model and others are sufficient enough to determine proper integration and leveraging of AI adoption. 

• Interviews and Surveys: Engaging with experts in AI adoption and governance, business transformation, and product development. 

• Literature Review: Analyzing existing research on digital implementation and disruptive technologies. 

Who (Qualification): 

Ideal candidates for this project are business professionals who are interested in AI, disruptive technologies, future industries, and business studies.  

Sponsor Introduction:

Ricoh is a leading provider of integrated digital services and print and imaging solutions designed to support digital transformation of workplaces, workspaces and optimize business performance. Headquartered in Tokyo, Ricoh’s global operation reaches customers in approximately 200 countries and regions, supported by cultivated knowledge, technologies, and organizational capabilities nurtured over its 85-year history. In the financial year ended March 2024, Ricoh Group had worldwide sales of approximately 16 billion USD.

This CIP project supports AI capabilities within Ricoh’s Intelligent Business Platform, aka IBP, which is a core part of the Digital Services Center within Ricoh USA.  

About the AI Pod:

The AI Pod is a group of developers within the Digital Services Center. We develop and support the document extraction and classification applications used in production within IBP. The infrastructure that supports these applications runs on Amazon Web Services. We utilize the latest developments in generative AI for document extraction and classification including advanced OCR, machine learning, and multimodal reasoning LLMs.

Project Description:  

At the beginning of the internship, onboarding meetings will be held, and based upon intern interest and fit, one of the projects below will be selected for them to work on.

• AI evaluation tooling development
  • Within our group, we have both ad hoc tools (pytest, python scripts, manual review) and sophisticated services (Sagemaker MLflow) that we use for evaluating the performance of our deployed AI solutions. With the rapid progress in AI, the way in which we evaluate these solutions must also constantly change.
  • In this project, interns will work on running models in our development environment, assessing limitations of our current evaluations, and contributing code to enhance our evaluation tooling and capabilities.
• New AI service assessment
  • With the rapid progress in AI, sometimes it is only weeks that separate advances in AI services. Indeed, new lower cost LLM models, higher performance LLM models, and LLM models with new capabilities become available to us that may be capable of saving significant development time, increasing performance,  or lowering deployed cost if deployed to production.
  • In this project, interns will work on learning and documenting the new services that we have in the queue for evaluation in our development environment, and contributing an MVP that we will use to determine if it merits moving to our production systems.

Objective: 

Contribute code and documentation to our Azure repos using the tools and standards all developers within our group follow.

Sponsor Introduction:

Learn more about the Tata Group here.

Project Description:  

This initiative aims to design, pilot, and validate a structured Human-in-the-Loop (HITL) Generative AI Ideation Framework that enhances team creativity, shortens ideation cycles, and improves quality of outputs. Inspired by Prof. Ethan Mollick and Prof. Christian Terwiesch’s work on AI as a co-creator, the framework will be deployed across select teams with measurable outcomes tied to creativity, efficiency, and scalability.

Deliverables: 

AI Ideation Framework v1.0: Process map, role definitions, prompts library, tools matrix

Pilot Run Reports: Use case, team feedback, metrics vs. baseline

GenAI Playbook: Documented methodology for broader rollout

Final Report & Recommendations: Metrics summary, lessons learned, scale plan