SenpAI: new consultants' AI mentor from entry to mastery

Inspiration

Societal Well-being as Our Anchor Point

AI is reshaping how humans work and live. Every day, new technologies, ideas, and services emerge—while traditional, stable ways of working rapidly fade away. We’ve entered an era of information explosion, where what you learned last year, month, or even last week may already be outdated. In this new reality, people are being both pushed and inspired to shift from repetitive, routine tasks toward highly creative, analytical, and human-centered work. We are dedicated to helping people become mentally and practically ready for this transformation — to learn faster, adapt more quickly, and embrace a more conscious, purpose-driven way of working. To support this shift, we provide an all-in-one solution that accelerates the essential phase of initial knowledge accumulation, encompassing:

1. Objective, reality-based knowledge — the practical “know-how” of daily operations and tools.
   
2. Subjective, abstract knowledge — higher-order thinking skills such as logical reasoning, negotiation, and problem-solving frameworks. Our solution incredibly simplifies the journey from “book to brain”, turning knowledge into living skills, and drives this evolution in a human-centered, approachable, and sustainable way.

Why Consulting Industry

Among all industries, consulting sits at the forefront of this transformation. Firms like Accenture act as global think tanks, delivering a full spectrum of services — from upstream executive-level strategy to downstream digital transformation execution. For consultants, constant learning is a way of life. They must continuously absorb new knowledge, stay ahead of emerging trends, and respond quickly to ever-changing business environments. Moreover, the project-based nature of consulting — rotating through new clients, roles, and teams every few months — demands not only adaptability and self-driven learning, but also exceptional mental agility and resilience.

Our Target Users: New Consultants

New graduates are among those who face the greatest challenges in adapting to the transformation of the consulting industry.
It’s not easy to “adapt fast,” “think logically,” and “stay confident” in the business world when one has only just stepped out of the academic environment. Many experience uncertainty, self-doubt, and isolation as they strive to perform.

The Core Problem

We identified three recurring pain points among new consultants:

1. Learning Retention Gap: Knowledge is poured in daily, but opportunities to review and practice are scarce — making it hard to internalize learning.
   
2. Communication Barrier: Unsure when or how to ask questions, many hesitate to speak up. Limited time and unclear frameworks make reporting and consultation difficult.
   
3. Confidence & Motivation Gap: Despite hard work, results don’t always follow. Feeling unnoticed and unappreciated, they lose confidence and motivation.

What it does

SenpAI acts as a warm, intelligent mentor that supports new consultants’ growth from entry to mastery through three specialized AI mentors:

1. Training Trace Senpai
   Reads through training materials, automatically generates understanding-check quizzes, and re-asks questions along the memory retention curve to ensure deep learning and lasting knowledge retention.
   
2. Communication & Writing Mentor Senpai
   
   • Speaking Practice: Offers feedback on how to frame questions and start conversations, helping users master structured and confident communication.
     
   • Writing Practice: Reviews weekly reports, emails, and inquiry messages, guiding users to build logical, structured writing. Goes beyond mere text correction to develop real-world communication and problem-solving ability essential in consulting practice.
     
3. Counseling Senpai Provides empathetic, story-based guidance to help users cope with common challenges in consulting life — such as pressure, human relationships, or failure experiences.
   Shares how “senpai who once struggled the same way” overcame similar obstacles, offering both comfort and actionable advice.
   Together, these three Senpai form an integrated system that listens, guides, and grows with each user.
   SenpAI enables anyone to learn self-drivenly, think proactively, take steady practice, and confidently embrace the constant change.

How we built it

1. Teamwork

We built SenpAI in an agile and fast-paced manner with a small, focused team — one backend developer and one frontend developer — working after hours.
Features were released in priority order, 1st prototype created within 1–2 weeks.
Development flow:
Brainstorming → Feature definition → Refinement & priority mapping (~1 man-day) → Parallel FE (~1 man-day) & BE development (~1 man-day) → Knowledge base setup (~0.5 man-day) → User testing.

2. System Architecture & Implementation

The entire system was built on AWS, adopting a service-oriented architecture where each function operates as an independent component.

Component	Main Services / Technologies	Implementation & Integration Details
Frontend	AWS Amplify, React, GitHub	Integrated GitHub with AWS Amplify to establish a CI/CD pipeline, enabling rapid deployment of source code updates.
Backend	AWS AgentCore Runtime	Deployed the developed AI agents to AWS AgentCore Runtime to serve as the backend’s core logic.
Authentication	AWS Cognito	Implemented user authentication including sign-up, sign-in, and session management.
Data Retrieval	S3, AWS Bedrock Knowledge Bases	Implemented RAG-based search functionality using data stored in S3 as the knowledge source.

Challenges we ran into

Challenge	Future Direction
Due to the specifications of AWS AgentCore Memory, there was no function to retrieve session lists, which made it difficult to design conversation history data.	Plan to research and learn best practices for efficient in-memory conversation management to achieve a more refined data design.
Without a project management tool, it was difficult to assign tasks and visualize progress in a multi-member development setup.	For future team development, we plan to introduce project management tools (e.g., Jira, Trello, GitHub Projects) to improve communication and task management efficiency.

Accomplishments that we're proud of

• We came up with a high-impact idea that combines strong business potential with positive social value.
  
• We built a brilliant cross-functional team of top performers across the company — each highly responsible, collaborative, and willing to contribute 2–5 extra hours per week beyond their busy project work.
  
• We successfully delivered an initial prototype within just few weeks, learning from zero to one and demonstrating exceptional adaptability and teamwork.

What we learned

Project Aspect

1. maintaining a shared understanding of the idea,
2. taking quick and decisive action, and
3. balancing limited time with the pursuit of quality. We embraced the mindset of completion over perfection, and learned to prioritize what is urgent and important to keep the project moving forward effectively.

Technical Aspect

one of the biggest takeaways was the opportunity to actively utilize AWS AI services and apply a modern development approach.

Category	Details	Insights
AI Agent	Deployment to AWS AgentCore Runtime	Learned that AI agents can be built with minimal development cost.
Development Tool	Amazon Q	Used it like pair programming and experienced a boost in development efficiency.

What's next for SenpAI: new consultants' AI mentor from entry to mastery

Our next steps are to:

1. Establish a scalable orchestration architecture to enable seamless coordination among SenpAI’s specialized agents.
   
2. Continuously release new features that deepen learning support and enhance user interaction.
   
3. Expand deployment — first across Accenture, then throughout the consulting industry, and eventually into other industries.
   
4. Evolve into a 2C application that supports individual growth and well-being across society.
   Through this gradual expansion, we aim to broaden our positive impact and ultimately make the world a better, more humanity-sparking place.

Built With

• amazon-web-services
• bedrockagentcore
• github
• langgraph
• next.js
• next.js-**platforms:**-aws-agentcore
• python
• typescript
• typescript-**frameworks:**-langgraph