With extensive experience in Generative AI, NLightN's mission is to enable businesses to effectively leverage the power of GenAI and stay ahead in a rapidly evolving technological landscape that has the potential to transform the way we do business.
Project 01 — Healthcare Analytics
Natural Language Processing for Healthcare Program Analytics
Uncovering actionable insights from unstructured data across high-volume healthcare initiatives
Developed an enterprise-grade NLP platform that processes patient records, lab and pharmacy records to deliver insights for one of India's largest public health initiatives.
Key Features
- Multi-modal data ingestion (PDFs, data from Healthcare application Database)
- Cross-program analytics using natural language queries
- Secure cloud deployment
Impact
- Empowered program leaders with clear, narrative-driven insights from complex clinical data, enabling faster decision-making
- Strengthened cross-program collaboration by surfacing hidden patterns in patient journeys and treatment efficacy
- Generated an initial level of interest among stakeholders in their journey towards GenAI adoption paving the way for future higher levels of adoption
Technologies
Python Flask
NextJS
AWS Cloud
Docker
MySQL
OpenAI
Claude
Pinecone
Project 02 — HR Assistant
PolicyBot: AI-Powered HR Assistant on PeopleSoft
Instant policy guidance for 1,500 employees across 40 Indian locations
Built a conversational AI assistant integrated into Oracle PeopleSoft, providing 24/7 access to HR policies, leave rules, and compliance updates.
Key Features
- Contextual Q&A with policy version tracking
- Voice mode input
- Multilingual support (English, Hindi, Tamil, Telugu)
- Audit logs for compliance reporting
Impact
- Transformed HR support into a proactive, always-on service, reducing employee frustration on policy misinterpretation
- Fostered a culture of self-service and compliance, freeing HR teams to focus on strategic people initiatives
- Delivered consistent, accurate guidance across regions and languages, reinforcing organizational fairness
Technologies
MS365 SSO
Azure
PeopleSoft
RAG architecture
Python Flask
NextJS
AWS Cloud
Docker
MySQL
OpenAI
Claude
Pinecone
Project 03 — Technology Leadership
Fractional CTO Leadership for National IT Foundation
Provided services as fractional CTO, on re-architecting for cost control, DevOps modernisation and delivery excellence for a foundation supporting 150+ tech grantees.
Key Features
- Cloud cost reduction roadmap
- CI/CD pipeline design with GitHub Actions
- Vendor evaluation and contract renegotiation
Impact
- Cut infrastructure costs by 41% ($1.2M annual savings)
- Improved deployment frequency from bi-weekly to 6x/day
- Enabled 99.99% uptime for mission-critical grant platforms
Technologies
AWS
OpenEdx
WordPress
Project 04 — Environmental Education
National Environmental Education Competition Portal
Engaging 1,500 schools and 10,000 students in climate action
Designed and deployed a scalable web portal with outreach workflows, partner portal, project submissions and evaluation workflows for India's largest school-level environmental competition.
Key Features
- Scalable Competition Portal: Built on cloud-native architecture to support 1,500+ schools and 10,000+ participants
- Outreach Workflows: helps stakeholder manage and monitor the reach out to partners, teachers and schools
- Secure Partner Portal: Role-based access for core team, partners, teachers, students and jury with dashboards and audit trails
- End-to-End Submission & Evaluation Engine: Drag-and-drop project uploads, AI-assisted plagiarism checks, multi-stage reviewer and evaluation workflows
Impact
- Provided the Core team an end to end view of the entire process and progress on a live basis
- Provided Partners better visibility on their school's progress
- Provided Schools a better shared team-based portal to move in steps
- Provided an easier workflow experience for the jury
Technologies
Python Django
Angular
Postgres
AWS
Docker
Project 05 — NPO Analytics
Cross-platform Analytics for India's Largest NPO
Engineered an AI analytics platform that delivers program performance insights to field workers and leadership through intuitive web dashboards and WhatsApp chatbots.
Key Features
- WhatsApp Business API for query-in-PDF analysis
- Natural language querying & dashboarding
- Natural language report generation
- Role-based access with granular permissions
Impact
- Enabling frontline workers to access insights via familiar tools like WhatsApp
- Shifted decision-making from intuition to evidence, strengthening program design and reporting
- Built organizational muscle in AI literacy, with staff now actively requesting new analytics features
Technologies
WhatsApp Business API
Python Flask
Angular
MySQL
Pinecone
OpenAI
Claude
Project 06 — Language Learning
Spoken English Coach with Moodle Integration
Personalized language learning for 6,000 rural scholarship students
Created an AI-driven mobile app that assesses pronunciation, grammar, and fluency, integrated with Moodle for seamless progress tracking and certification.
Key Features
- Real-time speech analysis with accent adaptation
- Adaptive lesson paths based on CEFR levels
- Activities & evaluations
- Teacher oversight dashboard
Impact
- Gave rural scholarship students a voice and confidence through personalized, judgment-free language practice
- Integrated seamlessly into existing academic workflows
- Demonstrated that AI can be a compassionate coach, not just a tech tool, in underserved communities
Technologies
Moodle LMS API
LAMP stack
OpenAI Whisper
Project 07 — Grant Processing
Document Processing for Major Grant Makers
Automating drafting and review of grant applications
Built an end-to-end AI pipeline to draft and review grant proposals, budgets, and impact reports for India's top philanthropic foundations.
Key Features
- Hybrid RAG based context engineering
- Auto-summarization and reviewer assignment
Impact
- Restored reviewer focus on impact, not paperwork, by automating tedious document handling
- Enhanced agility and pace in grant evaluation, building higher throughput
- Helped the foundation take an early step as an AI innovation leader
Technologies
Python Flask
NextJS
AWS Cloud
Docker
MySQL
OpenAI
Claude
Pinecone
Project 08 — L&D Platform
WhatsApp-Based L&D with AI Assessment
Micro-learning and instant evaluation for distributed workforces
Launched a WhatsApp-native learning platform delivering bite-sized training modules with AI-graded assessments.
Key Features
- Daily dose of section-by-section microlearning
- Adaptive quizzes
- WhatsApp integration for easy access
Impact
- Made learning accessible, bite-sized and fun via WhatsApp - fitting training into busy workdays
- Improved knowledge retention and application through instant, contextual feedback
- Proved that L&D can thrive outside traditional platforms, inspiring new approaches to workforce enablement
Technologies
WhatsApp Business API
Python Flask
NextJS
AWS Cloud
Docker
MySQL
OpenAI
Claude
Pinecone