Debasish Maji
Pioneering the Future of Generative AI & LLMs
Leading AI Agent Development at Microsoft | Building RAG-based systems that transform enterprise workflows | 9+ years driving innovation at Microsoft, Atlassian, and PhonePe
Leadership in AI Innovation
Transforming vision into reality through technical excellence and strategic leadership
AI Innovation Pioneer
Leading AI Agent Development for Personal Insights at Microsoft, implementing cutting-edge Agent-to-Agent communication and MCP protocol for enterprise-grade AI systems.
Revolutionizing how enterprises leverage AI for intelligent insights
Performance Excellence
Architected Rovo Agent at Atlassian, a RAG-based AI system that improved accuracy by 20% and developed user embedding systems enhancing search relevance by 12% for 10M+ monthly active users.
Measurable improvements in product performance and user experience
Scale Master
Built data pipelines at PhonePe serving 550M+ users and processing 6.71B annual transactions with sub-50ms latency using sophisticated microservices architecture.
Enterprise-scale systems handling massive user bases
Business Impact Driver
Enabled $1B+ in revenue through scalable usage tracking and billing solutions. Reduced code review time by 40% with AI-powered automation, directly impacting developer productivity.
Tangible business outcomes and operational efficiency gains
Driving AI Innovation
Transforming complex AI challenges into scalable, production-ready solutions
Building the Next Generation of AI Systems
Leading AI Agent development with cutting-edge LLM architectures and RAG systems
Advanced AI/ML Engineering
9+ years of software engineering with 4.5+ years focused on AI/ML and 3+ years pioneering Generative AI solutions. Deep expertise in LLM fine-tuning, quantization, RAG, and scalable AI infrastructure.
Research & Innovation
Conducting cutting-edge research on hallucination reduction in LLMs through architectural enhancements, hybrid search (dense & sparse retrieval), and self-supervised learning to improve accuracy and trustworthiness.
Beyond Technology
Leadership and innovation thrive on diverse experiences and fresh perspectives
Finding balance through exploration
Exploring nature and seeking inspiration
Embracing adventure and new perspectives
Life on the water
Drawing creativity from diverse experiences
Coastal inspiration
Bringing a holistic approach to AI leadership — combining technical excellence with global perspectives, creative thinking, and a commitment to continuous learning through diverse life experiences.
Learn AI by Building
A collaborative initiative where we learn AI together by building real-world projects. No lectures. No prerequisites. No fees.
Over the past few months, many people have reached out - not just to talk about AI, but to truly understand it, to learn by building, and to explore how it can solve real-world problems.
That genuine curiosity has been inspiring. So I decided to spend some of my personal time on weekends working with you on small, real-world AI projects.
Nothing formal - just a collaborative approach to learn together. No prerequisites - just bring your curiosity and a mindset to learn by experimenting.
Working on AI/ML and agent-based projects at Microsoft, Atlassian, and PhonePe - combined with learning from Coursera and DeepLearning.AI, researching AI hallucinations, and building research agents - has shaped how I think about intelligent systems. I genuinely believe we can build something meaningful together.
Learn by Building
No lectures, no theory dumps. Just hands-on projects that solve real-world problems
Collaborative Learning
Work in teams, share knowledge, and grow together with a global community
100% Free
No fees, no prerequisites, no barriers. Just bring your curiosity and willingness to learn
Real Projects
Build production-grade AI systems - RAG, agents, LLMs, and intelligent applications
Don't worry about it getting crowded - we can run multiple projects in parallel and form teams based on time zones. Share with friends or colleagues who might enjoy building something hands-on together!
Featured Projects
Production-scale AI systems driving measurable business impact
Microsoft
Workforce Insight Agent
AI-powered agent system providing intelligent insights for workforce management and optimization using LLMs and RAG architecture
15+
Data Sources
<2s
Query Speed
92%
Accuracy
Key Highlights
- •Multi-source data integration and synthesis
- •Natural language query processing
- •Real-time workforce analytics and recommendations
Microsoft
Agent Evaluation Framework
Comprehensive evaluation system for AI agents, measuring performance, reliability, and quality across multiple dimensions
1000+
Test Cases
20+
Metrics
95%
Coverage
Key Highlights
- •Automated evaluation pipeline
- •Multi-dimensional quality assessment
- •Performance benchmarking and regression detection
Atlassian
Rovo Agent - RAG-based AI System
Advanced Retrieval-Augmented Generation (RAG) AI system providing real-time intelligent assistance with enhanced accuracy through fine-tuning and quantization
+20%
Accuracy Boost
Low
Latency
Reduced
Hallucinations
Key Highlights
- •Optimized retrieval pipelines with FAISS and vector search
- •Reduced LLM hallucinations through enhanced retrieval efficiency
- •Integrated into Atlassian Intelligence ecosystem for automated insights
- •Enabled intelligent decision-making and AI-driven workflow interactions
Atlassian
AI-Powered Search in Atlassian
Next-generation intelligent search system leveraging LLMs and semantic understanding to improve content discovery across Atlassian products
550M+
Users Impacted
+12%
Relevance Boost
<200ms
Query Speed
Key Highlights
- •Semantic search with intent understanding
- •Cross-product search experience
- •Personalized ranking and recommendations
PhonePe
Real-Time Fraud Detection System
ML-powered fraud detection processing 100M+ transactions daily with 95% accuracy and sub-100ms latency
100M+
Transactions/Day
95%
Accuracy
<100ms
Latency
Key Highlights
- •Real-time feature engineering pipeline
- •Ensemble models with online learning
- •Reduced false positives by 30%
Professional Journey
9+ years of driving AI/ML innovation across industry-leading companies
Microsoft
- Leading a team to build an AI agent for Personal Insights within the PAX group
- Implementing sophisticated Agent-to-Agent communication for seamless coordination
- Leveraging MCP protocol for secure, standardized interactions across systems
- Designing scalable Multi-Agent framework for contextual and adaptive insights
- Ensuring enterprise integration with Microsoft's ecosystem for employee-facing use cases
Atlassian
- Designed and deployed the Rovo Agent, a RAG-based AI system, improving accuracy by 20%
- Developed user embedding system improving search relevance by 12% for 10M+ MAUs
- Led development of AI-powered Code Review Agent, reducing review time by 40%
- Architected real-time ETL-based usage tracking system achieving 10K RPS
- Enabled $1B+ in revenue through scalable usage tracking and billing solutions
PhonePe
- Engineered data pipelines serving 550M+ users and processing 6.71B annual transactions
- Developed near real-time recommendation systems with user embeddings (NCF + FAISS)
- Built MVAS platform driving 27% increase in merchant traction
- Achieved sub-50ms latency with robust microservices architecture
Zopper
- Led development of user account management and payment processing modules
- Scaled GMV from $15M to $25M through payment optimization
- Integrated 10+ payment/loan providers with optimized gateway switching
Technical Mastery
Comprehensive expertise across AI/ML, engineering, and senior leadership
AI/ML & Data Engineering
Software Engineering
Cloud & DevOps
Data Processing & Storage
Leadership & Management
Strategic & Business Skills
AI Insights
Sharing knowledge and perspectives on AI/ML trends, techniques, and best practices
The Future of RAG Systems: Beyond Simple Vector Search
Exploring advanced techniques in retrieval-augmented generation including hybrid search, re-ranking, and context compression for production systems.
Deploying LLMs at Scale: Lessons from 550M+ Users
Deep dive into infrastructure patterns, cost optimization, and reliability engineering for large-scale LLM deployments in production.
Building High-Performance AI Teams: A Practical Guide
Strategies for hiring, mentoring, and scaling AI engineering teams based on experience at Microsoft, Atlassian, and PhonePe.
Fine-tuning vs RAG vs Prompt Engineering: When to Use What
Comprehensive comparison of different LLM adaptation techniques with real-world examples and performance benchmarks.
Education & Credentials
Strong academic foundation and continuous learning through advanced certifications
Master of Computer Applications
Bachelor of Science
Professional Certifications
Let's Work Together
I'm available for consulting, freelance projects, teaching, mentorship, speaking engagements, and collaboration opportunities.
Book a time to discuss how we can work together to bring your AI vision to life.
Consulting
AI/ML strategy, GenAI implementation, and LLM architecture consulting for your enterprise
Freelance Projects
Custom RAG systems, AI agent development, and ML pipeline implementation
Teaching & Workshops
Technical training on Generative AI, LLMs, and enterprise ML systems
1:1 Mentorship
Career guidance, technical mentorship, and AI/ML expertise sharing
Podcast & Talks
Speaking engagements on AI innovation, leadership, and technology trends
Collaboration
Research partnerships, startup advisory, and strategic AI initiatives
Free initial consultation • Flexible scheduling • Global availability
Let's Connect
Interested in discussing AI/ML innovations, collaboration opportunities, or just want to connect?
Feel free to reach out through any of these channels.
© 2025 Debasish Maji. All rights reserved.