Skip to content

Gen AI & Data Engineering Workshop [Part-3]: AI Career Assistant for Double-Edge Growth

Featured Image

At Azilen, our GenAI and Data Engineering Workshop series has been about one clear idea: turning data into applied intelligence.

The third blog in this journey focused on something closer to each one of us – people, skills, and careers.

This time, our team presented GenAI-powered Career Assistant, a system that builds intelligent skill profiles and learning pathways using real-world data.

Where the Idea Originated?

Every organization gathers huge volumes of employee data, such as timesheets, project logs, resumes, certifications, and training records.

Yet, most of this information sits unused.

We wanted to see how this data could be turned into live intelligence that helps people and organizations understand skills dynamically – what exists, what’s emerging, and what’s next.

The Vision

To transform professional development by intelligently connecting people with peers, mentors, and personalized learning paths that accelerate career growth.

We see a future where AI becomes a career companion, which helps understand individual journeys, recognize emerging skills, and connect people to the right opportunities in a safe, authentic environment.

A big applause to the brilliant minds behind this initiative:

→ Karan Chokshi – Senior Software Engineer II

→ Nirmita Prajapati – Senior Software Engineer II

→ Rahul Gogia – Software Engineer

→ Dipali Rangpariya – Software Engineer

→ Karan Koradiya – Software Engineer

→ Mohit Kapadia – Software Engineer

→ Preksha Kharidia – UX Designer

→ Vedansh Kamdar – Associate Software Engineer

→ Rajesh Chaudhari – Associate Software Engineer

→ Rushabh Parikh – Associate Software Engineer

→ Shyama Shah – Associate Software Engineer

→ Manthan Bhavsar – Associate Software Engineer

→ Jaydeep Akhani – Product Owner

Here’s a glimpse from the workshop.

AI Career Assistant Workshop

The Solution: Career Assistant

Career Assistant is an AI-driven career intelligence system that creates a confidential space where employees can:

→ Track their professional growth based on real work data

→ Discover personalized learning opportunities

→ Connect with mentors and peers across the organization

At its core, it uses data from daily work, project management systems, and verified credentials to build a living, learning profile of each individual.

Generative AI
Want to Build Smarter Solutions with GenAI?
Explore our 👇

The Core Foundation of AI Career Assistant

It connects two real-world data streams that truly represent an individual’s capability:

Timesheet Data: Daily work details (tasks, job names, and project logs) reflecting skills in action.

CV Documents: Verified professional credentials, education history, and formally documented skills that establish baseline expertise

By analyzing both, the system builds a complete and evolving view of each employee’s skills.

The GenAI Processing Pipeline

During the workshop, our team walked through the data-to-intelligence journey. The pipeline orchestrates a multi-step analysis powered by GenAI:

1️⃣ Extract verified skills from CV documents

2️⃣ Scan timelogs to identify additional tools and technologies

3️⃣ Merge both datasets and remove duplicates

4️⃣ Normalize terminology for consistent labeling

5️⃣ Filter out generic tasks like “testing” or “PR comments”

6️⃣ Define the analysis period using recorded date ranges

The output is a precise skill graph, a living snapshot of each employee’s technical and non-technical capabilities, built automatically.

GenAI Prompt Architecture: Building the Skill Profile

The foundation of the system lies in its prompt-driven AI architecture, which ensures that every insight is explainable and structured.

The below prompt ensures that both documented expertise and real work performance are blended into one dynamic profile.

Prompt 1: Initial Skill Profile Generation

You are an AI assistant that builds an Initial Skill Profile for employees.Inputs:1. Timelogs JSON – employee_code, job_name, task_description, task_name, time_log_date2. CV document – professional details, education, and skillsTasks:- Analyze both sources together- Prioritize CV skills (ground truth)- Add new skills from timelogs if missing- Ignore generic activities- Deduplicate & normalize skills Name – Determine analysis period (earliest to latest date)- Output structured JSON with all profile components

Example Output: Employee Skill Profile

At the workshop, we demonstrated a live case of how this system generates profiles.

Employee Code: 1086

Name: Karan Koradiya

Analysis Period: Jan 2023 – Dec 2024

Technical Skills: Java, Spring Boot, React JS, AWS, Docker, Jenkins, Microservices, PostgreSQL, CI/CD, REST APIs, Kubernetes

Non-Technical Skills: Agile Development, Problem Solving, Client Communication, Code Review, Technical Documentation, Mentoring

Project Management Tool → Work Snippet Generation

The PM Tool integration is one of the most exciting modules in AI Career Assistant, built to turn real project conversations into structured intelligence.

1. Voice-Driven Project Updates

It begins with a speech-based interface where project managers share real-time updates about ongoing projects – what’s happening, who’s involved, and how the team is performing.

AI-generated questions guide these conversations, which makes it easy for PMs to narrate progress naturally while the system captures every detail through voice input.

2. Smart PM Summaries

Once the conversation ends, AI transcribes the dialogue and produces Smart PM Summaries – concise overviews that capture the project’s purpose, key contributors, milestones, and blockers.

Each summary comes with an entity correction layer, which allows PMs to verify employee names, project titles, and tools before storage.

This ensures that the data is clean, normalized, and ready for analytics.

3. Data Mapping in Neo4j

After validation, the summaries are pushed into Neo4j, our graph database.

Here, every project, person, and skill becomes a connected node.

This structure lets us visualize relationships between work, collaboration, and outcomes. In short, it forms the backbone for skill discovery and learning recommendations.

4. Automated Work Snippet Creation

From the graph, the system automatically generates Work Snippets – short, tweet-like reflections summarizing what each employee worked on, learned, or achieved.

Each snippet:

→ Pulls data from PM summaries, timesheets, and CVs

→ Adds context around tools, technologies, and responsibilities

→ Reads naturally, like a mini achievement post employees can review or refine

Employees can use AI-assisted refinement to polish their snippets before publishing them on their personal portal feed.

5. Enabling Contextual Reasoning with GraphRAG

Powering this intelligence layer is GraphRAG (Graph-based Retrieval Augmented Generation).

Graph RAG Implementation

Unlike plain RAG systems that read text sequentially, GraphRAG draws information from the Neo4j graph, understanding how entities relate – projects to people, people to skills, and skills to outcomes.

This means the AI doesn’t just recall data, it reasons through relationships and offers contextually accurate and explainable insights.

Want to learn more about RAG? Read the resources below:

➡️ When to Choose a RAG AI Agent

➡️ Agentic RAG Implementation

➡️ What Great RAG as a Service Looks Like

6. From Conversations to Insights

Together, these steps convert everyday PM updates into structured, interconnected knowledge.

It creates a living record of work and achievement, one that supports learning, performance tracking, and collaboration across teams.

Collaboration and Learning Intelligence

Our vision extended beyond profiling. The AI Career Assistant also enables collaboration and learning guidance.

We showcased three key GenAI prompt systems that bring this capability to life.

Prompt 2: Skill Matching

var prompt = $@”You are a senior technology mentor.

Given the user’s current skills: {string.Join(“, “, request.Skills)},

suggest 3-5 new relevant skills or technologies to learn next.

Respond ONLY with a JSON array of skill names”;

This prompt helps the system recommend next-skill priorities for each individual, tailored to their work history and expertise.

Prompt 3: Learning Roadmap Creation

varprompt=$@”Youareanexpertlearningpatharchitect.Createacomprehensive10-steplearningroadmap.LEARNERPROFILE:-TechnologytoLearn:{technology}-TargetDomain:{domain}-CurrentExperience:{experience}yearsExistingSkills:{string.Join(“,”,skills)}REQUIREMENTS:Createexactly10progressivelearningstepsthattakethelearnerfrombeginnertoadvancedin{technology}for{domain}applications.Eachstepmustinclude:-stepNumber:(1-10)-title:Clear,specifictitle-description:Detaileddescription(30-50words)-estimatedHours:Realistictimeestimate(5-10hoursperstep)-resources:Arrayof2-4specificlearningresources(books,courses,documentation,tutorials)RespondONLYinthisJSONformat:{{“”steps””:[{{“”stepNumber””:1,””title””:””Steptitle””,””description””:””Detaileddescription””,””estimatedHours””:40,””resources””:[“”Resource1″”,””Resource2″”],””completed””:false}}]}}”;

The output creates a personalized, machine-generated learning path that can plug directly into LMS or internal training portals.

Prompt 4: Peer Connection & Expert Matching

Peer Connection

varprompt=$@”YouareanAImatchmakingassistantforalearningplatform.Giventhefollowing{type}swhoareworkingwith{technology},rankthembywhowouldbethebestmatchesforsomeonelearning{technology}.Considertheirskills,experiencelevel,androle.Collaborators:{collaboratorsJson}ReturnONLYaJSONobjectwiththisstructure:{{“”rankedIds””:[arrayofcollaboratorIDsinorderfrombesttoworstmatch],””reasoning””:””Briefexplanationoftherankingcriteria””}}”

Expert Matching

varprompt=$@”Youareaseniortechnologycareeradvisor.Analyzethisdeveloper’sprofileandprovidepersonalizedlearningrecommendations.DEVELOPERPROFILE:-Name:{user.GetProperty(“name”).GetString()}-CurrentDomain:{currentDomain}-Experience:{experience}years-CurrentSkills:{string.Join(“,”,skills)}-LearningInterests:{string.Join(“,”,domains)}(domains),{string.Join(“,”,technologies)}(technologies)TASK:Basedontheirprofile,suggest:1.3-5complementarydomainstheyshouldexplore(includingtheirselectedones)2.4-6technologiestheyshouldlearn(includingtheirselectedones)3.Abriefreasoning(2-3sentences)explainingwhytheserecommendationsfittheircareertrajectoryRespondONLYinthisJSONformat:{{“”domains””:[“”domain1″”,””domain2″”,””domain3″”],””technologies””:[“”tech1″”,””tech2″”,””tech3″”,””tech4″”],””reasoning””:””Yourdetailedreasoninghereexplainingthestrategicfit.””}}”;

This mechanism helps identify peer learners (colleagues learning the same technology) and connects them instantly through Microsoft Teams integration.

Why AI Career Assistance Matters

For HR leaders and learning product companies, AI Career Assistance creates a live ecosystem of skill intelligence.

It helps enterprises:

✔️ Identify evolving competencies and learning gaps

✔️ Recommend relevant learning programs automatically

✔️ Support real-time workforce planning

✔️ Align individual growth with business needs

This bridges the long-standing gap between recorded skills and real capabilities.

Closing Thought

As we said during the workshop:

“Turning everyday work data into smart skill insights with GenAI.”

This workshop was another milestone in how we bring intelligence closer to human capability – practical, explainable, and ready for enterprise adoption.

AI Agents
Want Your Own AI Assistant to Simplify Workflows?
Explore our 👇

Top FAQs on AI Career Assistant

1. What exactly does AI Career Assistance do?

Think of it as your personal growth companion inside the organization. It keeps track of what you work on, what skills you use, and how you’re growing, all automatically. It connects your project data, CV, and timesheets to create a complete picture of your career progress. Then it suggests what to learn next, who to learn with, and how to get there faster.

2. How is this different from a regular HR or LMS system?

Most HR systems focus on records, such as job titles, appraisals, and static skill entries. Career Assistant focuses on reality, the actual skills you use every day. It continuously learns from your work and builds insights dynamically, which makes it much more personal and practical.

3. What is GraphRAG and why did you use it?

GraphRAG stands for Graph-based Retrieval Augmented Generation. It’s a GenAI technique we use to make our AI reason through relationships – between people, skills, and projects. So, when the AI answers a query like “Who are the best mentors for AWS?” it doesn’t just pull text, it understands context, connections, and relevance.

4. Can Career Assistant integrate with our existing LMS or HR systems?

Absolutely. The entire architecture is built with integrations in mind. The data it generates – skill graphs, learning paths, and performance insights – can sync with most enterprise LMS or HR platforms. It complements what you already have by adding intelligence on top of it.

5. What kind of organizations can benefit from this system?

Any company that wants to build a culture of continuous learning, whether you’re an enterprise, a SaaS provider, or an HRTech platform. It’s especially valuable for organizations that want to move from manual skill tracking to dynamic, AI-backed career growth systems.

Azilen Technologies
Team Azilen

Azilen Technologies is an Enterprise AI development company . The company collaborates with organizations to propel their AI development journey from idea to implementation and all the way to AI success. From data & AI to Generative AI & Agentic AI, and MLOps, Azilen engages with companies to build a competitive AI advantage with the right mix of technology skills, knowledge, and experience.  

Related Insights

GPT Mode
AziGPT - Azilen’s
Custom GPT Assistant.
Instant Answers. Smart Summaries.