AI Agent for Efficient Project Knowledge Management
Over 96 million knowledge workers in North America wasted nearly $2 trillion annually on repetitive tasks and information searches. Today, all these tasks can be easily and quickly automated.
This statistic once sparked the idea of KITRUM’s GetMeo project. As part of KITRUM’s internal innovation program, CEO Vlad Kytainyk initiated the GetMeo project to explore Al’s potential in solving executive productivity and knowledge management challenges. Custom Al Knowledge Agent, including all related intellectual property, was conceptualized and is solely owned by Vlad Kytainyk, CEO of KITRUM.
Story behind
Our initial exploration of this AI assistance concept began with the MyMeo project. MyMeo was initially conceived as a B2C tool aimed at individuals dealing with information overload, with a focus on AI Time Management. KITRUM CEO requested it and used it to manage his information effectively. Vlad initially explored this solution using Notion. However, this approach had limitations, as the Notion database became complex and challenging to manage.
To overcome these limitations and address a broader market need, we shifted our focus and developed GetMeo. GetMeo represents a transition from a B2C approach to a B2B solution, aiming to build a memory layer for companies. GetMeo reflects an ongoing effort to use AI to improve productivity and knowledge management, evolving from an initial personal tool concept to a more robust B2B solution.
Main challenges
The GetMeo project faced specific challenges, primarily related to the cutting-edge technology it employed.
Workflow
Our GetMeo development team was structured for efficiency. It consisted of a product manager, a full-stack developer, a backend developer, and a frontend developer. A key element of the GetMeo workflow was the close collaboration with a technology partner to ensure alignment with the latest AI technology updates we used.
GetMeo’s development spanned 3 months, including 1 month of ongoing collaboration with technology partners to integrate and adapt to AI technology advancements and 2 months of active development. The project is continuing — we are enhancing the feature list, which will take 2 more months.
The team adopted an Agile-inspired development approach, emphasizing rapid testing and validation. Project tasks were managed using Jira and Notion, while the team communicated through Slack, Telegram, emails, and regular meetings held 3-4 times per week.
Discovery phase
The GetMeo project's discovery phase focused on finding efficient solutions and identifying a viable market opportunity. Key aspects of this phase included:
Development phase
The development phase focused on creating a B2B solution for project knowledge management. The project aimed to develop AI knowledge agents to collect information from various sources into project-level "folders." This involved collecting data from sources such as messages, Slack threads, documents, and CRM information, aiming to consolidate the knowledge that could be reused for tasks like drafting emails and reports, answering questions, communicating internally and with clients, and updating CRM. A core aspect of GetMeo's development was integrating advanced AI technology for knowledge understanding and retrieval. To address the challenge of extracting and interpreting complex information from unstructured data, the GetMeo solution uses Retrieval-Augmented Generation (RAG) to make this information more accessible and usable. RAG, in general, is the technique of using external knowledge sources to enhance the capabilities of a large language model (LLM). In the GetMeo case, the specific type of RAG used is HybridRAG. This approach integrates GraphRAG (graph-based Retrieval-Augmented Generation) with VectorRAG (vector-based RAG) to leverage both strengths.
Implemented features
- Human-like chatbot
- Conversational interactions
- Collection of data and support in the form of emails
- Upload of documents in any format and quantity
- Support for YouTube videos
- Support for links from the internet
- All conversational and file exchange history saving
- Numerous other integrations with messengers such as Slack, note-takers such as Notion, etc.
Impact & Results
KITRUM acted as the technology enabler for the GetMeo product, applying its internal Al and software development capabilities. The company’s role was limited to development services and R&D acceleration.
This project falls under KITRUM’s policy allowing the CEO to pursue and lead founder-driven IP initiatives with support from internal resources. While the KITRUM team contributed engineering expertise, the product, vision, and legal rights belong to Vlad Kytainyk as the founder
and intellectual author.
By creating AI Knowledge Agents that collect and contextualize information, GetMeo aims to:
- Streamline information access, reducing the time employees spend searching for and organizing project-related data;
- Improve collaboration and decision-making by providing teams with quick contextual support;
- Enhance productivity by enabling employees to focus on higher-value tasks instead of being bogged down by information management.
GetMeo has the potential to significantly improve the quality of work-life by mitigating the negative impacts of information overload, such as stress and decreased productivity.
You may also like
Data Accuracy & GDPR Compliance with AI Healthcare Document Automation
How We Built a Flexible Vendor Management Platform Now Ready for White-Label Use
How Kitrum Simplified Tawkify’s Matchmaking Process
Subsea Leak Detection Management Software For Franatech
Big Data Migration From Hadoop to Databricks
Recommendation Engine Solution With Embedding-Based Retrieval Approach
Successful Migration and System Enhancement for Quantium
Embedded banking solution for FinTech Cross-Platform Startup
