About AI Innovation Challenge

The AI Innovation Challenge (AI-IC) is a flagship initiative of AI Center of Excellence, established by Department of Science & Technology and Gujarat Informatics Limited – Govt. of Gujarat in partnership with nasscom and Microsoft, to be a cornerstone of AI innovation, driving solution co-creation and impactful use cases to improve efficiency and quality in the public and private sectors.

AI Innovation Challenge is aimed at connecting India’s most promising AI and Deep-tech startups with real-world challenges from both Public & Private Sector organizations. The challenge provides a unique opportunity to showcase their cutting-edge solutions, collaborate with key stakeholders, and create meaningful impact in areas such as agriculture, healthcare, manufacturing, governance, and beyond.

Why Startups Should Participate ?
AI Innovation Challenge - II

Public Sector Use-Cases

Object Identification and Tagging from Images and Video

Trend, Seasonality, and Anomaly Detection on Time-Series and Tabular Data

Automated Blueprint Evaluation for Town Planning

Private Sector Use-Cases

Schneider Electric:
Enhancing customer service through the use of AI technology

Large Pharmaceutical Manufacturer:
Vision-Based Generative AI for Real-Time Image Analysis and Intelligent Recommendations

Epsilon Engineering:
AI led part identification through image comparison

Transworld:
Robotic Process Automation

Transworld:
Digitization of yard and gate management for containers

AI Innovation Challenge - I

Public Sector Use-Cases

Detection of Illegal Encroachment

Visitor Tracing using Facial Recognition

Bilingual OCR for English & Gujarati

Document Identification & Recognition

Identification of fake calls in ERSS 112

Integrated CCTV surveillance with FRS

Private Sector Use-Cases

Max Healthcare:
AI led Surgical Planning

RGCIRC:
Digitization of Legacy Medical documents

Charge Zone:
AI based predictive maintenance of EV Charging Stations

Tata Capital:
Employee Productivity using Chatbot Assistance

Use-Cases Challenges

Develop an AI-driven solution to automate encroachment detection on public or private land using image analytics. The solution aims to reduce manual efforts, improve detection accuracy, and provide a user-friendly dashboard for visualization. Targeted for urban development and municipal authorities, it should handle large-scale datasets efficiently while adhering to data privacy, achieving >95% accuracy on test datasets.

Design a secure facial recognition system for real-time visitor authentication and tracking in high-profile government buildings like Sachivalay. The solution aims to enhance security, streamline visitor management, and achieve 99% accuracy in identification. It must handle high visitor volumes securely while complying with data protection laws and delivering real-time analytics for administrative and security teams.

Develop a bi-lingual OCR system to accurately extract text in English and Gujarati from scanned documents and images, supporting diverse fonts and formats. The solution targets government departments, businesses, and academic institutions, aiming to digitize the process, reduce manual transcription efforts, and improve data accessibility. It must achieve >95% accuracy, handle low-quality scans and handwritten text, and comply with regional data privacy standards.

Develop an AI-driven solution to automatically identify, classify, and recognize document types like invoices, ID cards, and certificates while extracting key metadata fields. Targeting government agencies, banks, and businesses, the system aims to improve workflow efficiency with >90% accuracy. It must handle noisy, unstructured documents securely and comply with data privacy standards, significantly reducing manual processing efforts.

The use case focuses on leveraging AI to identify fake calls made to Emergency Response Support System (ERSS) 112. By analyzing call recordings and extracting Call Data Records (CDR), AI-powered speech recognition detects specific keywords, patterns, or anomalies in spoken words that may indicate a hoax. The system flags such calls in real-time and alerts authorities to prioritize genuine emergencies, improving response efficiency and reducing operational strain on emergency services.

The use case focus on a solution that integrates CCTV surveillance with Facial Recognition Systems (FRS) to enhance the identification of inmates, staff, and visitors using body-worn cameras. This system aims to improve security by enabling real-time monitoring and reducing the risk of unauthorized access. By leveraging FRS, the solution would address challenges like manual verification, delayed threat detection, and operational inefficiencies, creating a safer and more controlled environment.

Surgery for head and neck cancers poses significant challenges due to the proximity of critical structures. While clinical examination and radiological imaging provide valuable insights into the surgical landscape, unexpected complexities often arise during the procedure. By integrating AI-powered analysis of pre-operative imaging with immersive simulations of intra-operative conditions, surgeons can visualize surgical planes in relation to adjacent structures with enhanced accuracy. This approach not only improves pre-operative preparation but also reduces intra-operative uncertainties, leading to better outcomes. Additionally, this solution enables a training platform for surgical trainees, enabling them to practice and navigate challenging surgical scenarios in a controlled, risk-free environment.

Develop a GenAI-enabled chatbot to assist employees with a wide range of tasks, from content writing, document summarization, and email drafting to HR queries and customer service support. Designed for versatility, the chatbot facilitates learning and development, compliance, team collaboration, and automation of repetitive tasks. It will provide personalized assistance, streamline workflows, and enhance productivity while allowing for seamless integration of additional use cases in the future.

Develop an AI-driven predictive maintenance solution for EV charge point operators to transition from reactive to proactive maintenance practices. By analyzing real-time data from OCPP notifications, historical maintenance logs, and environmental factors, the system will predict potential failures, optimize maintenance schedules, and assign priority levels based on customer impact. It will include an AI-based alert system, actionable insights for technicians, and a centralized dashboard for monitoring charger health and maintenance efficiency, reducing downtime and improving operational efficiency.

Transforming legacy medical records into digital assets is often hampered by their non-digital format, complexity, and sensitive nature. Traditional manual processes are slow, error-prone, and resource-intensive. This can be overcome by an AI-led solution that enables automated extraction, classification, and digitization of medical documents with unparalleled accuracy and speed. The solution needs to have stringent privacy and security frameworks, ensuring the compliance with regulations, safeguarding patient confidentiality. The solution enhances operational efficiency, reduces costs, and ensures the long-term accessibility and integrity of critical medical data.

Government departments face challenges in manually analyzing visual data, leading to inefficiencies and inaccuracies. This AI-based solution automates object identification and tagging in images and videos, enhancing accuracy and reducing manual effort. Utilizing deep learning techniques like CNNs and transformers, the model supports real-time processing and large datasets. Applications include surveillance, infrastructure monitoring, and public safety. Deliverables include an AI model, documentation, and a demo. The solution improves decision-making, boosts efficiency, and sets a benchmark for innovation with hybrid AI approaches, ensuring scalable and accurate object recognition across various domains.

Government departments rely on structured and time-series data for decision-making, yet identifying trends, seasonality, and anomalies manually is inefficient and error-prone. This AI-powered solution automates data analysis to uncover hidden patterns, detect anomalies, and enhance forecasting accuracy. Using advanced AI/ML techniques such as LSTM, ARIMA, and ensemble learning, the model ensures real-time insights via a dashboard. The solution supports large datasets, missing values, and diverse data types while adhering to security regulations. By improving operational efficiency and resource allocation, this innovation drives data-driven policymaking, ensuring timely trend detection for better governance and strategic planning.

Manual blueprint evaluation for town planning is time-consuming and prone to errors, leading to compliance issues and project delays. This AI-powered solution automates blueprint analysis using advanced image processing and deep learning techniques to ensure adherence to town planning regulations. The system identifies non-compliance issues, reduces manual review time, and improves decision-making accuracy. Supporting multiple blueprint formats and high-resolution images, it provides real-time feedback through a visualization dashboard. By enhancing regulatory compliance, streamlining approvals, and improving efficiency for builders and town planning departments, this solution transforms the urban planning process with data-driven automation.

Our customer care agents manage numerous inquiries every day, ranging from simple questions to intricate issues, which require them to navigate various tools and coordinate with multiple teams, resulting in time-consuming processes and delays. We seek to implement agentic AI technology to automate this workflow, aiming to streamline information retrieval and communication, thus enabling our care agents to deliver prompt and accurate responses. It is crucial that this AI incorporates necessary checkpoints to ensure the reliability of the information conveyed to our customers, integrates seamlessly with existing tools, automates data extraction and analysis from emails, efficiently routes queries to relevant departments, provides real-time updates for agents, and maintains a user-friendly interface for both agents and customers. By achieving these objectives, we are poised to enhance our customer service operations, reduce response times, and significantly improve the overall customer experience.

We are looking for a Vision-Based Generative AI solution that utilizes advanced image analysis to provide real-time recommendations based on analytical reports. The solution should be capable of monitoring operator behavior in controlled environments, such as clean rooms, using ergonomic analysis to ensure compliance and efficiency. Additionally, it should detect and analyze anomalies like smoke patterns, providing real-time alerts and actionable insights. By integrating AI-driven image recognition with real-time reporting, the solution should enhance operational safety, process optimization, and regulatory adherence, ultimately improving decision-making and proactive intervention in industrial and high-precision environments.

For the variety of parts that flow through the manufacturing process, instant identification solution is desired comparing the scanned form of the part preferably on mobile camera with the design stored images and thus prompting its updation in ERP at the particular process stage.

Container Freight Stations (CFS) currently rely on traditional systems and processes involving manual intervention and extensive paperwork. Operations include repetitive and mundane task to be done on a daily basis. These conventional methods often lead to inefficiencies, delays, and increased operational costs. With the growing demand for more efficient and streamlined operations in the logistics sector, there is a significant opportunity to leverage emerging technologies like Robotic Process Automation (RPA).

Container Freight Stations (CFS) handle high volume container movements, which is currently manually handled at all our locations. 1. Tracking real-time container locations and vehicle movements becomes challenging. 2. The absence of real-time monitoring for yard space utilization can lead to suboptimal yard management and occasional congestion. 3. Management and utilization of equipment is also done manually. With an automated Gate and Yard Management Solution, we are looking forward to automated container handling, enhanced real-time visibility, improved resource allocation, and secure and compliant operations. The system should integrate with existing logistics infrastructure and provide actionable insights for better decision-making.

Program Structure

Launch of AI - Innovation Challenge

Registrations by Startups across the country

Startups’ Engagement Sessions with Stakeholders

Mentoring Sessions on Proposals by Technical & Business Experts

Final Evaluation by Jury Members & Winner Announcement

PoC Engagement with Winning Startups

Solution
Deployment

Eligibility Criteria for Participation Innovators and Mature Startups across the Nation, having Deep-tech capabilities and Market-ready Digital Solutions
  • Should have an annual turnover not exceeding Rs. 25 crore

  • Period of existence should not be exceeding 10 years from the Date of Incorporation

  • Should have the total manpower not more than 100 employees

Winning Innovators will Get
  • An opportunity to develop real-time solutions for a state government department through a paid proof of concept (POC).

  • 1 year membership to AI CoE Startup Accelerator Program

CoE will connect winner with the Use-case Owner Entity. The decision of Commercials to be taken by the Use-case Owners
Startups apply here

AI Innovation Challenge - I

AI Innovation Challenge - II

To know more about AI innovation challenge

Contact

AI Center of Excellence

Email : smartmanufacturing@nasscom.in or visit http://gujarat.coe-iot.com/