Hot Selling Street Finder – France
.-.
Goal : Help real estate agencies to spot the most favourable streets in France with dynamic reports
Goal : spot the most interesting streets for a real estate agent (number of sales, total turnover )
Data available : Open Data
Various GEOJSON files adding up to5 giga-byte and 25 million lines provided by the government listing 9 years of individual sales
Difficulties : all data had to be heavily cleaned as manyAnti notaries did not respect the standards while entering their sales into the system.
Result : success with this Power BI, with one clean single master data database gathering all sales in a place. Easy dynamic report generation allowing any real estate agent to sport the best streets in their zone in a few clicks .
Access : Corporate use – Not publicly available
Anti-Money Laundering / KY
.
.
Adress of most electricity-consuming residential buildings
.
.
Goal : Allow French city to spot the most electricity-consuming residential buildings of the city in order to prioritize improvements.
Data available : private energy company providing nationwide electic meter data as open data
Difficulties : heterogeneous ways of entering data by private employees
Success : Ability to type in a city and obtain immediately the exact address of the worst building in terms of electricity consumption
Access : Corporate use – Not publicly available..
Global coffee exporter historical race
.
.https://public.flourish.studio/visualisation/12217842/
.
Quick animation for education about the coffee producer race since 1961
Target : social networks
Site : https://public.flourish.studio/visualisation/12217842/
.COFF
Global coffee exporter historical analysis
.
.
Goal : as coffee is becoming a scarce commodity, allowing managers to study the rise of emergent coffee exporters allows for a better sourcing diversification, thus decrease the coffee refiners and coffee retailers sourcing risk.
Data available : World Economic Forum
Difficulties : Minimal data cleaning and transforming as data is well formatted at the source.
Success : very effective tool for corporate risk managers
Access : Corporate use. Not publicly available
.
Price estimation of real estate properties
/.
Goal : Use historical sales for a linear regression price estimation according to square meters and location, then ask user to input more personal criteria about the relevant property, to obtain a fair estimated value
Data available : historical sales as government open-data
Difficulties : data totaly unclean, very large data files.
Success : solid first self-price estimation of any real estate property in France, to be refined by the further visit of a real estate agent.
Internet site :
Goal : Allow French city to spot the most electricity-consuming residential buildings of the city in order to prioritize improvements.
Data available : private energy company providing nationwide electic meter data as open data
Difficulties : heterogeneous ways of entering data by private employees
Success : Ability to type in a city and obtain immediately the exact address of the worst building in terms of electricity consumption
Access : Corporate use. Not publicly available.
.
Anti-money laundering tool for financial institutions
.
.
Goal : gather all Sanction lists / Interpol Red Alert to allow the AML department to reduce time spent on searching though the different databases
Data available : public databases provided by different governments and entities
Difficulties : limited access ( Interpol ) and heterogeneous governmental databases
Result : One HTML / AJAX page written on PHP/MySQL allowing staff to simply write the few letters of a name and obtain immediate results. The application works on computers but also on smartphones.
Access : Publicly available at https://swissdove.ch
.
Guess the GPS location of properties described in real estate ads
.
.
Goal : automatically provide the exact GPS locations of properties according to the vocabulary and photographies used in real estate advertising on the French Riviera
Data available : government administrative open-data
Difficulties : availability of data, some real estate agents mistaking of lying when writing the property description, the massiveness number of properties.
Success : user became able to guess the exact GPS location of about 90% of the provided ads
Access : Publicly available : https://lapanthererose.com