Challenges for Innovators
These are real-world problems we'd like to see teams tackle. Each represents an opportunity to create meaningful impact with AI. Choose one that excites you, take it as extra validation to dive in, but you don't need to work on these ideas to participate in the hackathon.
You can bring your own problems.
All Challenges
2025 is shaping up to be the year of practical AI applications. We've put together a list of challenge areas that we think are especially promising for creating real-world impact through innovative solutions.
Can You Build Tech That's Not Just Functional—But Beautifully Designed?
Building something people actually enjoy using. Focus on creating beautiful, user-friendly experiences that solve real problems.
In today's world, building a working app or website is no longer the hard part. Thanks to AI tools, low-code platforms, and ready-to-use frameworks, anyone can launch a product quickly. But the real challenge? Building something people actually enjoy using. That's where great design makes all the difference.
Most Indian engineering students are great at coding—but often ignore design. The result? Apps that work, but are confusing, boring, or hard to use. The most successful tech companies—like Apple, Airbnb, and Stripe—didn't just build powerful tech. They created beautiful, user-friendly experiences.
Your Challenge
This hackathon isn't just about writing code—it's about building products that are usable, beautiful, and people-friendly.
Pick a real-world problem
Build a working solution (web/app/product)
Pay close attention to design, usability, and user experience
What to Focus On
Simple and intuitive interfaces
Clean layouts with thoughtful visual design
Features that solve a real user pain point
Smooth user flows (no clutter, no confusion)
Bonus: Use AI tools like Figma AI, Framer, or Uizard to boost your design process
Time Estimate
24-48 hours
Team Size
2-4 members
Impact
High - Create products that are not only smart but also a joy to use
Can Voice AI Fix Customer Support Calls?
Use Voice AI and LLMs to revolutionize customer service experiences and eliminate common pain points in support calls.
Think about your last call to a customer care number—maybe your bank, mobile service, food delivery, or hospital. You probably faced long wait times, confusing menu options, or had to repeat your issue multiple times. The experience hasn't changed much in decades, but with new Voice AI and Large Language Models (LLMs), we finally have the chance to fix this.
Today's AI voice models can understand natural language, respond with human-like tone and emotion, and hold realistic conversations. With over a trillion customer support calls happening each year worldwide, the opportunity to improve this space is massive—especially in India where phone-based support is still the primary way people reach businesses.
Your Challenge
Build a solution that uses Voice AI to improve customer service experiences and make customer calls faster, simpler, and more human.
A smart voice bot for a local business (e.g., pharmacy, clinic, delivery service)
An AI system that answers FAQs or resolves simple issues over a phone call
A voice assistant that understands regional Indian languages or accents
A hybrid system where AI handles the first 2 minutes and then hands off to a human if needed
What to Focus On
Natural language understanding
Human-like conversation flow
Regional language support
Seamless human handoff
Real-time voice processing
Time Estimate
30-48 hours
Team Size
3-4 members
Impact
Very High - Transform customer service for millions of users
Reducing Healthcare Inefficiencies with AI
Use AI to automate administrative tasks in healthcare, reducing paperwork burden and improving patient care efficiency.
India's healthcare system faces major challenges in how healthcare data is handled. From hospitals to clinics to insurance offices, a huge amount of time is spent on manual administrative work: filling out forms by hand, transferring data between systems, digging through patient records stored as PDFs, and manually updating information.
These inefficiencies slow down care, waste resources, and often lead to errors. Medical staff spend valuable time doing paperwork instead of treating patients. In the U.S., it's estimated that over $1 trillion is wasted on similar tasks. But today, LLMs and AI tools are making it possible to automate many of these tasks.
Your Challenge
Think about how you can use AI to reduce the paperwork burden in Indian healthcare and make the system more efficient.
Build a system that reads and extracts patient data from scanned prescriptions or reports
Help medical staff fill digital records more easily with voice input or chat
Move information from one hospital software system to another using AI
Create an assistant that supports health workers in rural areas to enter or retrieve data faster
What to Focus On
Document processing and OCR
Data extraction from unstructured sources
Voice-to-text for medical records
System integration and automation
Rural healthcare support tools
Time Estimate
36-48 hours
Team Size
3-4 members
Impact
Very High - Make Indian healthcare more efficient and patient-friendly
Build an AI Tutor for Every Indian Student
Create personalized AI tutors that adapt to individual learning styles and help students understand complex topics.
Engineering students in India face real struggles: huge class sizes with little personal attention, one-size-fits-all lectures, pressure to memorize without understanding, and outdated learning tools. Most online courses still deliver the same content to every student, regardless of their pace or learning style.
For decades, computer scientists have dreamed of using technology to revolutionize learning. That dream is finally possible with today's AI. With the latest LLMs and multimodal AI tools, we can now build AI tutors that adapt to each student's pace, break down tough topics step by step, and answer follow-up questions like a real teacher would.
Your Challenge
Build an AI-powered learning tool that can help Indian students understand hard topics in engineering, science, or any other subject.
Create visual explanations for complex ideas
Let students interact with concepts (e.g., manipulate 3D objects or ask questions)
Support multiple Indian languages to make learning inclusive
Help students learn at their own pace with a personalized AI tutor
What to Focus On
Personalized learning algorithms
Interactive visualizations
Multi-language support
Adaptive content delivery
Student progress tracking
Time Estimate
24-36 hours
Team Size
2-4 members
Impact
Very High - Transform education for millions of Indian students
Can You Reimagine Education in India with AI?
Reinvent how students learn and teachers teach in India—using AI to make education more personalized, accessible, and impactful.
India’s education system serves over 260 million students and 9 million teachers. Yet, traditional teaching methods—lecture-based learning, rote memorization, one-size-fits-all materials—remain dominant. Now, with AI, we can build tools that adapt to student needs, support teachers, and improve learning outcomes, especially for underserved regions.
Despite the size of India’s education sector, many students—especially in rural or low-income communities—are left behind. AI offers a chance to create scalable, personalized solutions that actually meet diverse student needs. This is about shaping the future of how India learns.
Your Challenge
Build an AI-powered solution that transforms any part of India's education system—especially where scale, personalization, and access are challenges.
Identify a core problem in the Indian education system
Design and build an AI-driven solution (platform/tool/agent)
Focus on impact, scale, and local usability (e.g., languages, connectivity)
What to Focus On
AI tutors that teach in regional languages
Teacher tools to grade or create content faster
Voice or visual learning for low-connectivity areas
Smart assistants for students to summarize, quiz, and translate content
Bonus: Use GPT, Gemini, or multimodal tools to create rich learning experiences
Time Estimate
36-48 hours
Team Size
2-4 members
Impact
Very High - Help millions of Indian students and teachers
Can You Build India's Personal Finance Copilot with AI?
Use AI to help individuals across India manage money better—through smart, unbiased, and accessible financial tools.
Most Indians rely on friends, YouTube, or guesswork to make financial decisions. Expert help is scarce, expensive, and often biased. With LLMs and APIs, we can now build accessible AI finance tools that provide personalized advice to anyone with a smartphone.
Millions across India make poor financial decisions due to lack of guidance. A simple, AI-powered copilot can help improve financial literacy and long-term wellbeing—especially in Tier-2/3 cities.
Your Challenge
Create an AI-powered tool that acts as a personalized financial guide for users in India—covering saving, investing, budgeting, loans, and taxes.
Identify a core financial pain point (e.g., budgeting, taxes, investing)
Build an AI advisor that gives simple, customized suggestions
Ensure support for Indian users—languages, tax rules, local context
What to Focus On
AI tools for tax planning or investment advice
Budget planners with voice/vernacular interfaces
Debt management and credit improvement tools
Education tools that explain finance in plain language
Bonus: Build for Bharat—focus on non-metro audiences
Time Estimate
24-48 hours
Team Size
2-4 members
Impact
High - Empower people to make smarter money decisions and plan their financial future with confidence
Can You Build Vertical AI Agents That Automate Real Work?
Design AI agents that don't just talk—but think, act, and automate real-world work in specific Indian industries.
India is full of manual, rule-based jobs in tax, healthcare, compliance, and service sectors. Vertical AI agents can replace these with intelligent systems that deeply understand the domain, integrate with real tools, and deliver end-to-end automation.
This is the next evolution of SaaS—vertical AI that can reason, automate, and replace repetitive work. It’s a massive opportunity to improve productivity and build AI-native businesses in India.
Your Challenge
Build a domain-specific AI agent that automates complex work—like tax filing, compliance checks, support calls, or data processing—at Indian scale.
Pick a real-world workflow that’s manual and repetitive
Design a domain-specific AI agent that can fully or partially automate it
Include real integrations (PDFs, voice, Excel, APIs) and Indian localization
What to Focus On
AI agents for tax, billing, support, logistics, compliance
Support for Indian languages and legal/tax frameworks
Integration with real-world data (documents, APIs, forms)
End-to-end workflows, not just chatbot front-ends
Bonus: Deploy in a simulation or real-world mock scenario
Time Estimate
36-48 hours
Team Size
2-4 members
Impact
Very High - Automate massive amounts of manual work across India's industries with vertical AI solutions
Can You Build an AI Assistant That Transforms Doctor-Patient Workflows?
Design AI tools for doctors and patients that streamline prescription creation and medicine reminders—powered by voice and intelligent automation.
Doctors in India are often overburdened and short on time, while patients struggle with prescription adherence and instructions. With voice AI and smart reminders, we can simplify both ends of the healthcare experience.
Paper-based prescriptions are error-prone and time-consuming. On the other side, many patients skip doses or misunderstand instructions. With voice-enabled AI assistants, we can create a smoother, safer healthcare loop for both doctors and patients.
Your Challenge
Build an AI assistant that listens to doctors’ voice commands and generates prescriptions, plus a patient-facing app that reminds and validates medicine intake.
Build a voice-driven prescription assistant for doctors
Create a companion app for patients to track medicine schedules
Add proof-of-intake or confirmation mechanisms (e.g., camera or voice prompts)
What to Focus On
Voice-to-text for medical use cases
Prescription generation and digitization
Automated reminders and alerts
Patient engagement and compliance tracking
Bonus: Support for multiple Indian languages and accents
Time Estimate
24-36 hours
Team Size
2-4 members
Impact
High - Improve healthcare efficiency and medication adherence
Can AI Help Us Make Climate-Friendly Daily Decisions?
Build a mobile-based AI system that alerts users before climate-damaging actions and crowdsources weather data from individuals.
Despite growing awareness, most people don’t realize when their everyday actions harm the planet. Meanwhile, India lacks dense, real-time weather data. AI and crowdsourcing can help solve both problems at scale.
Behavior change is critical to fighting climate change, and hyperlocal data is key to better climate models. A mobile AI assistant can help with both—by raising awareness and gathering insights directly from citizens.
Your Challenge
Create a dual-purpose app: one that nudges users to act climate-consciously in real time, and another that collects hyperlocal weather data from users (with consent) to improve forecasting.
Build an app that detects and alerts users before high-emission actions (e.g., unnecessary vehicle use, AC usage)
Create a consent-based crowdsourcing system for users to upload local weather data
Incentivize participation and ensure user-friendly data input (e.g., voice or image-based)
What to Focus On
Real-time behavioral nudges using AI
Crowdsourced sensor/data input
Location and privacy-aware app design
Climate education via micro-interactions
Bonus: Integrate with environmental APIs or IoT devices
Time Estimate
24-36 hours
Team Size
2-4 members
Impact
Very High - Help individuals take climate-positive actions while improving national weather data
Can You Build a Browser Lens That Filters the Internet for You?
Create a browser extension that stays with your query, intelligently highlighting or summarizing relevant content across webpages.
Search engines today show you AI answers and links—but once you click, you're on your own. Users are stuck digging through pages. An AI overlay that guides your browsing experience could massively reduce this friction.
Whether comparing product specs, researching a topic, or fact-checking across sites, people spend too much time scanning for answers. A persistent, context-aware browser layer can make digital exploration smarter and more efficient.
Your Challenge
Design a persistent browser assistant that remembers your query and helps you find the most relevant information on every page—especially useful for shopping, research, or comparisons.
Build a browser extension that tracks user queries
Overlay relevant highlights, summaries, or insights as users browse webpages
Allow dynamic follow-ups or zooming into specific content on the page
What to Focus On
Browser extension development
Contextual NLP for summarization and relevance
Real-time page augmentation
Support for shopping, learning, or research scenarios
Bonus: Multi-tab memory and query refinement
Time Estimate
36-48 hours
Team Size
2-4 members
Impact
High - Make the internet less overwhelming and more efficient for everyday users
Can You Build a Monitor That Detects Elderly Emergencies Without Being Asked?
Build a smart wearable that detects emergencies for elderly users based on abnormal motion, silence, or possible falls—without needing them to press a button.
Elderly users often wear bracelets or pendants that have emergency buttons. But if they fall unconscious or are unable to reach the button, help may not arrive in time. Intelligent monitoring can solve this.
India has over 140 million elderly people. Safety technologies must move from reactive to proactive to truly protect them—especially those living alone or with medical conditions.
Your Challenge
Current devices rely on elderly people pressing buttons during emergencies, but what if they can’t? Your task is to build a smarter, proactive alert system using wearables.
Design a wearable or mobile app that detects lack of motion or sudden falls
Trigger a soft prompt ('Are you okay?') and escalate to emergency call if no response
Integrate features like GPS tracking, heart rate, and ambient sensors (if available)
What to Focus On
Fall detection using accelerometers/gyroscopes
Emergency protocols and smart escalation
Low power wearable integration
Bonus: Combine with voice response and visual cues
Time Estimate
24-36 hours
Team Size
2-4 members
Impact
High - Could save lives in situations where the elderly can't call for help themselves
Can You Build an AI Assistant That Helps People With Dementia Recall Things?
Create a mobile app for elderly people with dementia that listens to memory-related queries and helps them recall names, faces, and context using available data.
Dementia can impair short-term and associative memory. What if we built a kind of AI-powered memory jogger—something like Google, but tuned for personalized, everyday recall?
Families and caretakers of dementia patients struggle to help with constant memory lapses. A patient-friendly assistant could preserve dignity and increase independence.
Your Challenge
Design an assistant that can intelligently fill memory gaps when a user says things like 'What’s Kumud’s son’s name again?', using cues from contacts, photos, and call logs.
Build a voice-enabled assistant for casual memory queries
Use context from contacts, messages, call logs, and photos to infer and answer
Support multi-modal prompts (e.g., 'Show me his photo', or 'When did we last talk?')
What to Focus On
Voice assistant for elder-friendly UX
Personal knowledge graph from local data
Multimodal recall (text, images, timeline)
Bonus: Privacy and caregiver integration
Time Estimate
24-36 hours
Team Size
2-4 members
Impact
Very High - Empowers dementia patients and reduces burden on caregivers
Can You Build an Executive Assistant That Preps You Before Every Meeting?
Create a tool that gives busy professionals a crisp briefing before each meeting—pulling relevant insights from past emails, CRM, calendar, and the web.
Executives jump from meeting to meeting. Context switching is hard, and prep often slips through. A smart briefing assistant can turn chaos into clarity.
From client names to pitch decks, the info is scattered across inboxes and CRMs. An intelligent summarizer could save hours every week and improve decisions.
Your Challenge
Build a system that auto-generates a pre-meeting briefing using the invite info and context from various sources like email, documents, and web data.
Pull meeting info from calendar
Correlate with past emails, CRM entries, documents, and relevant external data
Create a short, readable meeting brief (e.g., summary, last touchpoint, open tasks)
What to Focus On
Email and calendar API integration
Contextual summarization with LLMs
Secure data handling
Bonus: Slack/Teams notifications before the meeting
Time Estimate
12-24 hours
Team Size
1-3 members
Impact
Moderate to High - Save time and raise the bar on meeting effectiveness
Can You Detect Diseases From X-rays Faster Than a Doctor?
Build an AI tool to analyze X-rays and detect diseases like tuberculosis, pneumonia, and cancer—supporting faster, more accurate diagnoses.
Radiologists are in short supply, especially in rural areas. Many diseases go undetected until too late. AI models can now match or exceed human performance in interpreting medical images, making early intervention possible.
Early detection saves lives. With overburdened healthcare systems, AI can augment radiologists and improve care access across India.
Your Challenge
Design a diagnostic AI system that can detect anomalies in X-rays and assist doctors in identifying diseases like lung cancer, pneumonia, and tuberculosis.
Use open datasets or dummy medical image sets
Train a model to detect multiple diseases in chest or breast X-rays
Build an interface that gives interpretable visual outputs
What to Focus On
Medical imaging (X-rays, scans)
Disease detection (TB, pneumonia, cancer)
AI interpretability
Low-resource model deployment
Time Estimate
36-48 hours
Team Size
3-4 members
Impact
Very High - Help doctors detect diseases earlier and save lives
Can You Help Farmers Get Real-Time Advice in Their Language?
Design a multilingual advisory system for Indian farmers that gives real-time crop, weather, and market advice using geospatial data.
Smallholder farmers lack access to timely and relevant crop advice. Language, digital illiteracy, and scattered information make it worse. AI + Bhashini + maps can change that.
Improved decisions mean better yields, income stability, and less crop loss. This directly impacts India's food security and rural economy.
Your Challenge
Build an app that combines satellite insights and crop models with voice-based farmer support in local languages.
Pull geospatial data for weather, pests, crop health
Generate actionable advice in Indian languages
Include voice/chat interface for farmer interaction
What to Focus On
Multimodal language support
Crop health analytics and mapping
Voice assistants for rural users
Farmer-centric UX design
Time Estimate
24-36 hours
Team Size
2-4 members
Impact
High - Improve income and resilience for millions of farmers
Can You Build an AI System That Sees Retinal Issues Before You Do?
Create a solution that scans retinal images to detect diabetic retinopathy, cataracts, or conditions needing keratoplasty.
India faces a rising burden of diabetes-related blindness. With fewer ophthalmologists per capita, AI can aid early detection and support surgical planning.
Blindness is often preventable if caught early. AI can democratize diagnostics in eye care, especially in remote areas.
Your Challenge
Develop an AI tool that screens for eye conditions using retinal images, reducing blindness and enabling timely treatment.
Analyze retinal image datasets
Detect early signs of diabetic retinopathy or cataracts
Include explainable outputs to aid doctors
What to Focus On
Computer vision in healthcare
Explainable AI for diagnostics
Blindness prevention
Surgical planning assistance
Time Estimate
24-36 hours
Team Size
2-4 members
Impact
High - Reduce avoidable blindness and improve surgical outcomes
Can AI Be the Bridge Between Citizens and Government?
Build a conversational agent in Indian languages that helps citizens access public services and register grievances.
Many Indians—especially in rural areas—struggle to access government services due to language and literacy barriers. AI can simplify access and boost inclusion.
Bridging the citizen-government gap can unlock welfare, support legal empowerment, and improve service delivery.
Your Challenge
Create a multi-language AI assistant that provides information on schemes, files complaints, and navigates government processes.
Design a chatbot or voice agent powered by Bhashini
Pull data on government schemes or legal FAQs
Enable complaint registration or service discovery
What to Focus On
Voice/text chat in Indian languages
Legal and public service information
Multimodal accessibility (text, voice, image)
Grievance management integration
Time Estimate
24-36 hours
Team Size
2-4 members
Impact
Very High - Empower citizens and strengthen digital governance
Can You Help Kids With Learning Disabilities Learn Better?
Build tools to detect dyslexia and other learning disabilities early—and provide gamified, adaptive learning tools to support them.
SLDs like dyslexia are common but rarely diagnosed in India. Early support is crucial for self-esteem and success.
1 in 5 students may be affected, yet most go undiagnosed. Inclusive education needs AI to scale timely support.
Your Challenge
Create an AI-based system to detect and support students with learning disabilities through early identification, gamified lessons, and adaptive content.
Build a diagnostic screening tool for SLDs in Indian languages
Create gamified, accessible learning content
Support different modalities (audio, text, visuals)
What to Focus On
Early detection of dyslexia/dyscalculia
Gamified and adaptive learning tools
Local language and neurodiverse UX
Inclusive education support
Time Estimate
24-36 hours
Team Size
2-4 members
Impact
Very High - Enable inclusive education for millions of students
Can You Predict Disasters Before They Strike?
Build AI systems that forecast extreme climate events like floods, cyclones, and heatwaves—giving early warnings and saving lives.
India suffers from recurring climate disasters. AI can give authorities and citizens precious time to act—minimizing loss of life and property.
Early warnings are the frontline of disaster defense. With climate events growing more erratic, scalable forecasting is now a necessity.
Your Challenge
Use AI to build early-warning systems for extreme weather. Pull from open datasets and simulate disaster alerts.
Model extreme events using weather datasets
Predict floods, heatwaves, or air pollution spikes
Simulate alert messages and action prompts
What to Focus On
Disaster forecasting with AI
Satellite data and weather analytics
Crisis communication automation
Mapping high-risk zones
Time Estimate
36-48 hours
Team Size
3-4 members
Impact
Very High - Save lives and resources in vulnerable areas
Can You Auto-Convert Figma Designs Into Production-Ready React Native Code?
Build an AI-powered tool that reads Figma designs and outputs clean, modular React Native code—accelerating app development and preserving design fidelity.
Developers often spend hours translating static Figma files into mobile code—manually implementing layout, styles, and interactions. This process is slow and error-prone. With AI, we can turn design into code instantly and accurately.
Speed and consistency are critical in modern app development. Automating the design-to-code pipeline can reduce friction, eliminate human error, and help teams ship better apps, faster.
Your Challenge
Develop a tool that bridges the gap between designers and developers by automatically converting UI/UX designs from Figma into working React Native code.
Parse Figma design files or UX screenshots to extract layout and UI elements
Generate clean, modular React Native code with components like View, Text, Image, TouchableOpacity
Preserve design properties: layout, color, fonts, spacing, responsiveness
Ensure that the code is readable, maintainable, and follows best practices
Bonus: Support reusable design systems or component libraries (e.g., map Figma buttons to custom components)
What to Focus On
AI-based layout parsing
Code generation (React Native)
Preservation of design structure and hierarchy
Design system/component mapping
Bonus: Support for responsiveness and modular architecture
Time Estimate
36-48 hours
Team Size
2-4 members
Impact
High - Automate the design-to-development pipeline and boost product velocity