CONTACT US

Full Name required

Valid Email required

Message required

Thanks! We'll get back to you soon!

Aviation

Aircraft on ground costs billions, ZASTI helps reduce downtime of aircraft.

Accelerated by:

Aircraft Positioning
Aircraft Positioning

Let's build low-cost Aircraft position map for Airports

How
Aircraft Proximity Warning
Aircraft Proximity Warning

Let's predict collision by analyzing proximity of aircraft in Airports

How
Claims Prediction
Claims Prediction

Let's predict claims ratio and claims profile

How
Flight Delay Prediction
Flight Delay Prediction

Let's Predict delays for a Flight, Airlines or Airport!

How
In-Flight Surveillance
In-Flight Surveillance

Let's identify the Blacklisted or No Fly Listed Passengers in Airports or In-flight Cabin

How
Predictive Maintenance
Predictive Maintenance

Let's predict cycles left until failure for Turbo Fan Engine

How

back to aviation

Aircraft Positioning

Let's build low-cost Aircraft position map for Airports

Business Challenge

During landing procedure of an aircraft, the pilot turns off the engine due to which the transponder device gets turned off. This makes it difficult for the teams at the airport to help pilots navigate to the nearest available exit.

ZASTI Solution

ZASTI AI platform processes the images captured from the cameras installed at the runway. It identifies the aircraft which has just landed thereby coordinating the route of the aircraft successfully with pilot and AIS team at the airport.

back to aviation

Aircraft Proximity Warning

Let's predict collision by analyzing proximity of aircraft in Airports

Business Challenge

Rare incidents recently at Israel and Toronto airports wherein tail ends of two aircraft have collided on tarmac causing heavy damages.

ZASTI Solution

ZASTI AI platform processes the data obtained from the cameras at the airports or the proximity sensors mounted on the aircraft to calculate the proximity between the aircraft and potential threat. Integrated with an alert system to alert the cockpit or security in the event of the object coming closer to the aircraft beyond a permissible limit. Thereby averting any potential collision or threat risk.

back to aviation

Claims Prediction

Let's predict claims ratio and claims profile

Business Challenge

In 2014 over 368 Million, Passengers were affected due to flight delay issues. Insurance companies would like to know in advance if a passenger were to make a claim. This information is instrumental for the insurance companies to be prepared with cash flows.

ZASTI Solution

ZASTI AI platform processes over 5 years of claims data by passengers and learns patterns about people. Based on the learning, it continuously identifies profiles of customer and categorizes as Low, Medium and High Risk. As a result, insurance companies can easily predict eligibility of customer and help underwriters offer customized premiums, offer, discounts based on the profile.

back to aviation

Flight Delay Prediction

Let's Predict delays for a Flight, Airlines or Airport!

Business Challenge

In 2014 - 4.6 million flight were delayed affecting 368 million passengers. Aircraft on the Ground costs the Aviation industry £5 billion every year. Inaccurate models result in poor prediction of flight delay.

ZASTI Solution

ZASTI AI platform processed 5 years of data from airline carriers, insurers, weather, and maintenance to predict delay of Airline or Route for the next quarter or year. With 20% more accuracy than the current model. This also helped the model to predict claims ratio. The model was further tuned to predict individual flight delays. This helped insurance companies to identify risk routes, airlines, or periods and structure new products.

back to aviation

In Flight Surveillance

Let's identify the Blacklisted or No Fly Listed Passengers in Airports or In-flight Cabin

Business Challenge

No-fly list contains about 81,000 names and about 28,000 records are present in terrorist screening database. It is very difficult to track the movement of these people by just tracking their names as there have been incidents when innocent people bearing the same name have been barred from taking the flights.

ZASTI Solution

ZASTI A-eye uses state-of-the-art AI Deep Learning algorithms that have been developed by training millions of real-life images for Suspect Recognition and tracking Intruders or Persons of Interest at the airport or inside flight cabin. Integrated with existing alert systems to Alert security team or management based on the identification of a person from the No-Fly or Blacklisted Passengers List.

back to aviation

Predictive Maintenance

Let's predict cycles left until failure for Turbo Fan Engine

Business Challenge

According to international safety statistics, there are about 25 incidents a year involving a jet engine failing either in flight or on the ground. The most frequent causes of engines breaking up being ingestion of object, bird strikes, and mechanical failures Unplanned maintenance results in a loss for airlines and insurers.

ZASTI Solution

ZASTI AI platform's deep learning model processed 8,400,000 anonymous data points from 400 flights and built a custom deep learning model to predict remaining useful life of the component by analysing specific failure patterns. ZASTI solution can predict with a higher accuracy in comparison to many other leading AI platforms. The model was also able to identify top 5 influencers which helped the services and procurement team to re-align maintenance plans.

Healthcare

ZASTI helps companies, hospitals, doctors, and patients in early diagnosis of diseases

Incubated by:

IIT MADRAS

Blindness
Blindness

Let's predict Glaucoma 5 times earlier than traditional methods

How
Colon cancer
Colon cancer

Let's predict the survival rate of colorectal cancer patients

How
Predicting Depression
Predicting Depression

Let's track the health of a patient suffering from depression

How

back to healthcare

Blindness

Let's predict Glaucoma 5 times earlier than traditional methods

Business Challenge

It is estimated that 4.5 million people are diagnosed with Glaucoma globally. Of these people, 50% in developed countries and 90% in underdeveloped countries, are unaware of their condition. It takes 7-10 days for detection of glaucoma.

ZASTI Solution

ZASTI created a new architecture for image segmentation using adversarial network and with extensive experimental evaluations showed that it outperforms several state of the art techniques for the task of joint optic disc and cup segmentation. The result being ZASTI can detect Glaucoma in 5 seconds from Retinal Fundus Image.

back to healthcare

Colon Cancer

Let's predict the survival rate of colorectal cancer patients

Business Challenge

According to WHO, cancer is a leading cause of death worldwide, accounting for 8.8 million deaths in 2015. Colorectal cancer is the third most common types of cancer, leading to 774 000 deaths. Its early detection is difficult as it affects humans after the age of 50.

ZASTI Solution

Zasti AI platform performs retrospective analysis of patterns, including evolving similarities for unifying patient information across studies and developing homologs, prototypical patterns, exceptions and nomograms. Predictive analytics is used by AI platform to find early stage prognostic markers as well as personalized models for specific traditional and alternative end point predictions.

back to healthcare

PREDICTING DEPRESSION

Let's track the health of a patient suffering from depression

Business Challenge

300 million people around the world have depression, according to the World Health Organization and it’s estimated that 15 percent of the adult population will experience depression at some point in their lifetime. Tracking the health of patients suffering from depression at all times is difficult for doctors and family members.

ZASTI Solution

ZASTI AI platform uses data collected from Fitbit (with heart rate sensor) and psychometric questions to profile patients suffering from depression and tracks the status of their health in real time . The AI identifies anomalies in the pattern over time and flags family alert about the patient. This way movements resembling recurring depression are accurately tracked with an accuracy of over 85 % and alerts can be raised to provide medical help to the patient in need on time.

Insurance

ZASTI helps insurance companies and their clients, across varied industries such as marine, healthcare, aviation, in predicting risks and improving efficiency

Accelerated by:

Commercial Property Risk
Commercial Property Risk

Let's predict fire peril risk for a building and factors that cause it

How
Customer Risk Profile
Customer Risk Profile

Let's predict the eligibility of customers on the basis of risk profiles

How
Flight delay
Flight delay

Let's predict flight delays and accurate cash flows

How
Medical Insurance Fraud
Medical Insurance Fraud

Let's identify fraudulent medical claims and predict such frauds in future

How
Predicting Depression
Predicting Depression

Let's track the health of a patient suffering from depression

How
Vehicle Insurance Fraud
Vehicle Insurance Fraud

Let's identify the extent of damage to a vehicle and help insurer detect fraud

How
Marine Insurance Claims
Marine Insurance Claims

Let's identify fraudulent claims by ship time charters

How
Ship captain Risk Profile
Ship captain Risk Profile

Let's identify the risk associated with a ship captain based on various risk profiles

How

back to insurance

Commercial Property Risk

Let's predict fire peril risk for a building and factors that cause it

Business Challenge

Of the domestic property insurance claims in the UK, 15% are due to fire. Hence, knowing beforehand the fire risk of a building can help insurance companies charge extra premiums.

ZASTI Solution

ZASTI AI processed 1 Million+ commercial property insurance records with 0.2 Billion Data Points and 299 variables. It identified 15 variables that had a key impact on the outcome. The model was built and trained based on these data points. It predicted Loss Ratio for building profiles based on various policy parameters. This gives the insurance company the ability to profile buildings based on past data and public datasets such as weather, building classification and so on, and to predict risk level calculation based on policy and client parameters.

back to insurance

Customer Risk Profile

Let's predict the eligibility of customers on the basis of risk profiles

Business Challenge

Only 40% of U.S. households own individual life insurance. An insurance company takes an average of 30 days to identify risk and eligibility of existing and new customers. The challenge here is to automate the process to predict eligibility based on risk profiles of customers.

ZASTI Solution

ZASTI AI platform uses existing data sets and learns patterns about people. Based on the learning, it continuously identifies profiles of customer and categorizes as Low, Medium and High Risk. As a result, insurance companies can easily predict eligibility of customer and help underwriters offer customized premiums, offer, discounts based on the profile.

back to insurance

Flight delay

Let's predict flight delays and accurate cash flows

Business Challenge

An aircraft on ground costs £5 billion per year. An aircraft delayed for more than 3 hours costs the insurance company £600 per person. Insurance companies today have poor delay and inaccurate cash flow prediction methodologies.

ZASTI Solution

ZASTI AI platform processed 5 years of data from airline carriers, insurers, weather, and maintenance to predict delay of Airline or Route for the next quarter or year. With 20% more accuracy than the current model. This also helped the model to predict claims ratio. The model was further tuned to predict individual flight delays. This helped insurance companies to identify risk routes, airlines, or periods and structure new products.

back to insurance

Medical Insurance Fraud

Let's identify fraudulent medical claims and predict such frauds in future

Business Challenge

Annual medical fraud losses in the UK could be £193 billion. Identification of fraudulent claims and their validation has been challenging for insurance companies due to use of inaccurate models and methodologies.

ZASTI Solution

ZASTI AI platform uses Adversarial Neural Network to analyse insured’s medical claim images. The AI identifies if the image is valid by first checking if the image matches the profile of a customer and secondly by verifying if the same image was used previously (by same customer or another) and immediately validates a claim and presents insights to the insurer.

back to insurance

Predicting Depression

Let's track the health of a patient suffering from depression

Business Challenge

300 million people around the world have depression, according to the World Health Organization and it’s estimated that 15 percent of the adult population will experience depression at some point in their lifetime. Tracking the health of patients suffering from depression at all times is difficult for doctors and family members.

ZASTI Solution

ZASTI AI platform uses data collected from Fitbit (with heart rate sensor) and psychometric questions to profile patients suffering from depression and tracks the status of their health in real time . The AI identifies anomalies in the pattern over time and flags family alert about the patient. This way movements resembling recurring depression are accurately tracked with an accuracy of over 85 % and alerts can be raised to provide medical help to the patient in need on time

back to insurance

Vehicle Insurance Fraud

Let's identify the extent of damage to a vehicle and help insurer detect fraud

Business Challenge

Motor insurance fraud is estimated to cost the UK insurance market over £1 billion annually. Identification and validation of fraud been a challenge as it is difficult to determine whether the damage is under legal purview or the images of accident vehicle were used before but until now.

ZASTI Solution

ZASTI AI platform uses Convolutional Neural Network to identify damages based on videos and images. Once the model is trained for a particular country, vehicle, models, and features, it can identify the damage and extent of damage with a good amount of accuracy. This is how insurers can detect fraud.

back to insurance

Marine Insurance Claims

Let's identify fraudulent claims by ship time charters

Business Challenge

According to a contract between ship owners and the charterer, the ship has to perform according to specified fuel and speed conditions. Time charterers frequently bring claims against owners for underperformance. Underperformance claims often go hand in hand with claims for “overconsumption.” Insurance companies find it difficult to settle these claims legitimately as existing methodologies are not accurate.

ZASTI Solution

ZASTI AI platform provides mathematical validations using existing resources such as captain's noon report and data from GPS. This can be done by installing and analyzing images from the camera to monitor sea & control rooms and existing AIS data to track the vessel using deep learning techniques. The discrepancy between the captain’s logs and claims can be easily corroborated through above data and analysis.

back to insurance

Ship captain Risk Profile

Let's identify the risk associated with a ship captain based on various risk profiles

Business Challenge

Ship captain negligence, undeclared cargo, overweight containers, poor maintenance and sailing in restricted areas are hard to track and cost $3-6 billion annually. Ship owners find it difficult to measure the risk associated with a ship captain while hiring him and insurance companies find it hard to approximate the premium.

ZASTI Solution

ZASTI AI platform interprets large sets of data, such as vessel statistics, local weather, GPS, claims information of captain and combines this with historical information to reveal the behaviours that correlate to claims. Using this information, ship owners can identify risks associated with captains, and insurers, reinsurers can offer customized marine policies.

Compliance

ZASTI helps business and their clients by predicting potential risks and ensuring safety

Compliance Level
Compliance Level

Let's predict Health & Safety Compliance

How
Face recognition
Face recognition

Let's enhance Security using Facial Recognition

How
Orphaned Object Recognition
Orphaned Object Recognition

Let's Identify an Orphaned Object at a location

How
Risk Event Recognition
Risk Event Recognition

Let's predict untoward risks in a place/event

How
Suspect Recognition
Suspect Recognition

Let's Identify a suspect / person of interest from a sea of people

How

back to Compliance

Compliance Level

Let's predict Health and Safety Compliance

Business Challenge

Audits are carried out to ensure industrial workplaces such as construction sites, mechanical shop floors, health facilities, petroleum refineries, etc. are kept free from potential hazards that may lead to injury, illness, near miss, property damage or adverse environmental impact. Continuously monitoring safety compliance of a facility even after the audit can be a tedious task when done manually.

ZASTI Solution

ZASTI A-eye platform developed by training millions of real-life images uses advanced image processing algorithm based on Deep Learning. The model was trained on more than 40,000 objects such as hardhats, attires, special uniforms, people movement, ladders, scaffolding, etc. This helps the AI to automatically identify compliance adoption using video feeds from CCTV.ZASTI A-eye can predict compliance breaches based on the past and present data. Integrated with existing alert systems to Alert security team or management based on compliance breach. Hence ensuring 24 x 7 x 360 o compliance using existing infrastructure.

back to Compliance

Face recognition

Let's enhance Security using Facial Recognition

Business Challenge

There is an increased need for enhanced surveillance and monitoring of workplaces. But the infrastructure costs for installing Biometric and Retina scan systems at all access points ,can run up to a few millions, making it an unviable option for the organizations.

ZASTI Solution

ZASTI A-eye solution can be integrated with the existing authentication infrastructure of employee access cards, but augmented with another layer of security. ZASTI A-eye platform uses state-of-the-art AI Deep Learning algorithms developed by training millions of real-life images for Facial Recognition. The model can identify a known and unknown person based on existing images in the database. Thereby ensuring surveillance and monitoring in Organisations at minimum infrastructural changes and cost.

back to Compliance

Suspicious/ Orphaned Object Recognition

Let's Identify an Orphaned Object at a location

Business Challenge

Video Blindness sets in human CCTV operators after staring at the screen for 20-40 min. About 45 % of Screen activity is missed after 12 min and up to 95 % of the screen activity is missed by a human after 45 min. Thereby making it difficult for humans to keep track of orphaned or suspicious objects.

ZASTI Solution

ZASTI A-eye can Identify Risk Objects such as an Orphan bag developed by training the model using millions of object images. The AI platform uses state-of-the-art AI Deep Learning algorithms for Orphaned Objects Tracking .Integrated with existing alert systems to Alert security team or management based on suspicious orphaned object identification.

back to Compliance

Risk Event Recognition

Let's predict untoward risks in a place/event

Business Challenge

More than 300 million plus surveillance cameras worldwide and only 6% of the are monitored. In current Surveillance scenario, less than 35 % risk events are detected using human monitoring of CCTV footage.

ZASTI Solution

ZASTI A-eye can Identify and notify Risk Events such as a person falling down or a fight been developed by training millions of real life images. The AI platform uses state-of-the-art AI Deep Learning algorithms for Risk Event Recognition .Integrated with existing alert systems to Alert security team or management based on event identification.

back to Compliance

Suspect Recognition

Let's Identify a suspect / person of interest from a sea of people

Business Challenge

No-fly list contains about 81,000 names and about 28,000 records are present in terrorist screening database. It is very difficult to track the movement of these people by just tracking their names as there have been incidents when innocent people bearing the same name have been barred from taking the flights.

ZASTI Solution

ZASTI A-eye uses state-of-the-art AI Deep Learning algorithms that have been developed by training millions of real-life images for Suspect Recognition and tracking Intruders or Persons of Interest at the airport or inside flight cabin. Integrated with existing alert systems to Alert security team or management based on the identification of a person from the No-Fly List or Terrorist Screening Database.

Aviation

Aircraft on ground costs billions, ZASTI helps reduce downtime of aircraft

How?

Healthcare

ZASTI helps companies, hospitals, doctors, and patients in early diagnosis of diseases

How?

Insurance

ZASTI helps insurance companies and their clients, across varied industries such as marine, healthcare, aviation, in predicting risks and improving efficiency

How?

Compliance

ZASTI helps business and their clients by predicting potential risks and ensuring safety

How?

Verticals

AVIATION

HEALTHCARE

INSURANCE

COMPLIANCE

take a look at us

ZASTI's Team

A diverse team of Scientists, Strategists, Technologists, and Risk Analysts with international experience. Strategic expertise from an advisory board and mentors from diverse verticals, highly successful in multiple ventures and research

Krish

Chief Executive Officer

Dr John domenech

Chief Operating Officer
USA

Anna lisa

Mentor

Ram

Chief Strategy Officer

Raj

Chief Technology Officer

Vandana

VP HR and Compliance

Dr. Mohan

Chief Scientist

Herman De Latte

Advisor

Peter Tobin

Advisor

Raj Sinhal

Advisor

Jaimin

VP Marketing

Pete Blake

Mentor

Want to know more

Contact

London

Suite 1, 3rd Floor, 11 - 12 St. James Square, London,
SW1Y 4LB.

India

IIT-Madras Research Park,
Kanagam Road, Tharamani,
Chennai 600 113.


Information: reach@zasti.ai
Careers: join@zasti.ai

Drop Us A line

Full Name required

Valid Email required

Message required

Thanks! We'll get back to you soon!

Thank you for your message. If urgent, Contact reach@zasti.ai

oops...something went wrong

please try again later