Some of the exciting projects we've been working on:
Integrating data science into an early-stage product
About Farm Me Up
This startup aimed to provide funds to French farmers via a smart scoring system powered by Artificial Intelligence and backed up with Blockchain transactions system.
How did Dathappy help Farm Me Up to validate their business idea?
We mapped the processes and interactions between the key stakeholders to define how the scoring should work
We identified what kind of data is essential to create a smart credit scoring system for the farmers
We built personas and asked the relevant questions to validate the hypothesis and business model
“Shaping our startup idea, we knew Data will be important. But it could be challenging to understand if data sources are difficult to gather or what we need exactly to build smart systems. The strategic workshops gave us a valuable asset to build strong foundations for our next ideas and developments”
- Marie-Laure, Farm Me Up CEO
Results
The strategic workshops helped the CEO, the team and farmers to test the initial idea and develop a different business model that has answers to the French farmers’ problems better. Understanding the potential of Artificial Intelligence opened up new perspectives for the future of Farm Me Up
Understand the profiles of your users with data
About
Woomentum is a mobile platform for women to learn about the mechanics of building a business, solicit advice from mentors, get coaching, or raise funds.
Within only a few days, Dathappy helped us with Data visualizations for our Marketing team while showing us the power of AI on text classification.
- Mouna, Woomentum CEO
Objectives
-
Use the available data for market segmentation to create more relevant marketing content for the customers
-
Use Machine Learning to categorize the platform users into different profiles by automatic text classification
What happened?
Prepare
Quite often, an email list or database has multiple and incorrect entries just as a result of human error and users figuring out how to use your system. Cleaning and consolidating the information lead to more relevant conclusions and actions.
Build
Based on the text input entered by the user for the “Why I joined” question, using Natural Language Processing and Machine Learning, we categorized the users into different profiles and created a dashboard where the users can be filtered real-time based on different criteria.
Results
Apart from the insights themselves, this Proof of Concept was an eye-opener on the power of Data Science and its methodology. Having access on a daily basis to a dashboard helped to better understand and target the users while Machine Learning helped to reproduce tasks previously realized by humans.
Analyzing the link between volunteering and leadership
In 2018, Conjunct Consulting conducted a case study, funded by Tote Board to understand the link between volunteering and leadership development. Specifically, the research sought to understand:
-
What motivates youth to volunteer with nonprofits and the impact of volunteering on leadership development.
-
The benefits and challenges of youth volunteering for nonprofits.
-
The societal benefits of engaging youth volunteers with nonprofits.
How was the research conducted?
Both quantitative and qualitative research methods were undertaken, including a survey with Singapore youth aged 15-34 years old as well as interviews with key stakeholders and subject matter experts.
Dathappy supported the survey, bringing advanced Data analytics to extract more insights like the profiling of respondents.
What did we do?
We analysed the previous surveys to better orient the interviews
We reviewed the new survey questionnaire from data analytics viewpoint
We analysed and visualized the results of the new survey in a report
Findings
We have identified the 2 biggest retention challenges of the NPO sector.
Responding to the challenges, the NPOs identified a number of strategies to improve volunteer engagement.
Based on our survey data, we have created some sample volunteer profiles to showcase the main characteristics of the youth leaders/volunteers. This can be used for future recruitment and retention.
Impact
Yoann is a highly-skilled and talented data scientist who has a good understanding of the research process. He contributed to both the design and analysis of the survey and applied advanced data analytics to uncover additional insights from the data, which he presented in a creative and innovative way. I would highly recommend Dathappy and Yoann for future projects.
- Dr. Brigid Trenerry, Research Fellow, Singapore University of Technology and Design
About Conjunct Consulting
A Singapore-based nonprofit organization that offers pro bono management consulting services to other nonprofits and social enterprises in the areas of financial sustainability, human capital development, operations management, impact assessment, and strategic planning.
About Tote Board
As a broad-based grantmaker, Tote Board identifies social needs and challenges and through grants, it aims to achieve three strategic outcomes – providing equitable opportunities for vulnerable groups, building a resilient community, and enhancing quality of life for Singaporeans.
Understand the visitors and revenue streams better with real-time automated dashboards
About Mega Adventure
This amusement park’s mission is to give people the opportunity to challenge themselves through adventure. They operate Asia's #1 zipline, treetop ropes course, a free-fall simulator and a bungee-assisted trampoline.
During the internship, skills such as data exploration, data cleaning and analysis would be developed. And with the opportunity to tackle real-world data and business tasks, the internship would prove to be a challenging and rewarding experience
- Jae Chew, Business and Computer Science graduate
Objectives
-
Create a dynamic, visual dashboard to compare visitors by profiles by period and find actionable insights
-
Create a dashboard to monitor revenue by period and find actionable insights
What happened?
To deliver this project, we empowered 3 first-year MSC students, within a short internship to learn more about Data Science and its applications. Together, we went to the client attraction park to better understand the operations and challenges. Based on the needs, we recommended outcomes, a scientific methodology and tools to ease the life of the Mega Adventure staff and to help them make better informed decisions.
Results
The new real-time dashboards about the visitors and the revenue helped Mega Adventure to:
-
Understand their customer profiles
-
Compare the rate of return and popularity of each attraction by customer segments
-
Plan their digital marketing accordingly
Evaluate the impact and success of a government initiative
While most social service organisations recognize the need for volunteers and effective volunteer management, not every social service organisation has the ability or the resources to use and manage volunteers efficiently or effectively.
Between 2016-2018, NCSS launched the Volunteer Manager Funding Scheme (VMFS) to test the effect of a dedicated volunteer manager on social service organizations. NCSS selected 10 organizations and asked Conjunct Consulting and Dathappy to analyse the impact and success of the program in a report.
What did we do?
Design metrics and survey
1
2
3
4
Prepare and clean the data
Visualization
Extract insights
Findings
In surveying volunteers it became clear that when an organisation has a plan in place for volunteer management that includes orientation, training, recognition, and feedback processes, retention rates are higher. Volunteers are also happier and will passionately recommend the organisation to others looking to give to a social service organisation.
Impact of the program
Having a dedicated volunteer manager for two years has created positive change for each of the 10 participating organisations—from improving service quality to having happier, more satisfied volunteers.
Dathappy collaborated with us on a social impact report and was a highly effective and outstanding contributor. The team was able to analyze complicated sets of data and draw useful insights, which resulted in a lasting impact on the sector.
- Kwok Jia Chuan (CEO of Conjunct Consulting)
About Conjunct Consulting
A Singapore-based nonprofit organization that offers pro bono management consulting services to other nonprofits and social enterprises in the areas of financial sustainability, human capital development, operations management, impact assessment, and strategic planning.
About NCSS
National Council of Social Services is the umbrella body for over 450 social service organizations in Singapore. Its mission is to provide leadership and direction in enhancing the capabilities and capacity of its members, advocating for social service needs and strengthening strategic partnerships, for an effective social service ecosystem.
When data science meets Sustainable Development Goals in Sudan
About UNDP
The United Nations Development Programme (UNDP) is the United Nations' global development network. It advocates for change and connects countries to knowledge, experience and resources to help people build a better life for themselves.
What did you like the most about working with Dathappy?
“The real-life client interfacing”
- Anita Okoh (project team member)
Objectives
While working towards the Sustainable Development Goals (SDGs), UNDP engaged Dathappy to create an easily manageable and searchable catalog of the potential solutions to the SDGs. The catalog would help to:
-
Map the existing solutions in the world and make them available to the masses in an easily searchable way so the local communities can find the relevant solutions to their problems and adopt them,
-
Raise awareness towards those SDGs where the solutions are missing.
What did we do?
We recruited a team of 3 data enthusiasts for this mission: Emeka, Anita and Lucky from Nigeria.
We designed a reusable survey to map the solutions, priorities, challenges about the SDGs.
We translated the Arabic text to English and cleaned the data using drop-testing and indefinite rows.
We created a dashboard using Google Data Studio, then Power BI, and infographics in Canva.
Results
UNDP is now able to:
-
understand the geographical focus areas and reach of the solutions,
-
have an overview of the projects, their status and aims,
-
see which SDGs are prioritiezed by the innovators,
-
compare different ecosystems, agencies, validation statuses.
When Cyber Security leverages on AI to identify vulnerabilities
About GuardRails
GuardRails is a startup that is building the future of application security. Its mission is to empower modern development teams to find, fix and prevent security bugs in their code, without being overwhelmed or needing external security expertise.
Objectives
While application security has been around for quite some time, one of the key problems is that the existing solutions are creating a lot of false alerts. GuardRails decided to cut down the false alert noise through applying Machine Learning.
GuardRails asked Dathappy to help them enhance the current alarms classification of potentially dangerous scripts by developing Machine Learning models.
What did we do?
Data exploration & preparation
The team of four Nigerian data enthusiasts started the project by understanding and exploring the existing data structure. We recreated the database to have more info on each vulnerability and then visualized it in Google Studio.
Machine Learning Test
Using historical data (GuardRails historical decisions/classification is based on traditional rules), we trained our first model to enhance and identify the decision rules by Machine Learning algorithms
Improve the accuracy of the ML algorithm
In the remaining time, we focused on boosting the overall model accuracy and on optimizing the script to reduce runtime. We incorporated feedback from users to continuously improve the model and automated the data pre-processing to ease the deployment.
Results
We continuously delivered insights and tools with 2 weeks sprints to make sure we are always aligned with the business needs of GuardRails.
By the end of the project, we implemented a ML model that is able to identify the false positive alerts and help GuardRails’s clients to ship apps faster and more securely.
What the team said
“I wanted experience working on a real project and that's what I got. I also liked the support and Yoann's patience with us throughout the project.”
— Gigi Kenneth, project team member
Improve the efficiency of an Organic & Fair Trade Cambodian pepper farm
About Fair Farms
The Fair Farms founders’ dream was to create a community living in harmony in a small paradise in Cambodia while producing a top-quality Kampot pepper and respecting nature.
Objectives
-
To replace the manual data collection process with a digital offline tool that is easy to handle for the local operations manager, and to prevent manual entry errors;
-
To help the decision-making by having the most important harvest KPIs, parcel and worker productivity indicators surfaced in a real-time dashboard,
-
To be able to monitor the stock and forecast the quantity of the black/yellow/red/green pepper by parcel
2. Build interactive dashboards for 2 different business users
We defined two final users (personas) of the dashboards and their goals in order to design the dashboards according to their needs:
Operations Manager:
-
Wants to get visibility on the progress of the harvest, ideally real-time to be able to ensure the different phases of the harvest cycle are going according to the expectations
Business Manager (= the founders):
-
Want to be able to foresee the total volume for the entire harvest production in order to negotiate sales with clients and prospects according to the products that will be in stock
What did we do?
1. Introduced a new data collection process
As the farm doesn’t have a network, we needed a tool that works offline and can be synchronized online later by the local manager. We decided to use ZOHO that is integrated with Google Sheet, therefore made it easy to consolidate and visualize the data. We also set up automatic email alerts for unusually high/low metrics that indicate incorrect manual entries.
Contact us to make data-based business decisions!
Results
This project was an excellent example of how data can help any type of organization to operate more efficiently, even if the organization is at a remote location, without Internet and electricity.
We transformed Fair Farms into a data-informed organization that is able to make smarter decisions daily by exploiting data in all fields of their operations from tracking and forecasting their harvest to tracing back weak quality peppers to their parcels and batches.