Cody an Intelligent Coder Assistant

Currently, digitalization and automation are taking center stage around the world. One of the basic tenets of this trend is the shortage of software coders worldwide. This is because software coding is a daunting task even for STEM students when using traditional text-based coding method. Especially for young adults who is our primary target group. For each region around the world (America, Europe, Asia), there is a dominant economic region and a supporting economic region regardless of industries type. In Western Europe, is supported by the Eastern Europe region. In America, North America region is supported by the South America region. In Asia, North Asia (predominately China, Japan, Korea etc) the emerging supporting region is ASEAN. Our immediate Target Market is defined as Malaysia, Thailand, Vietnam, Indonesia and Philippines). Among the 3 regions (America, Europe, Asia), Asia is anticipated to be new source of global growth, whether in industrial or digital transformation. Within the context of ASEAN, the durable relationships and cultural cohesiveness has found new synergies for this economic bloc over the past few decades. Singapore situated centrally in this region is an ideal launchpad for the Asia region and already serving as catalytic agent in many different areas.      

An important observation from recent research shows that university non-engineering students using tools that does not require text-based coding, can perform better than engineering students. In the digital domain, software coding is a fundamental skill to be acquired. However, the teaching and learning methodology is still very traditional (long and difficult learning process) and skewed towards STEM tertiary educated cohort. Vast majorities of youth are deprived of this learning opportunity.

Therefore, re-engaging youth learners in digital skills, needs a disruptive paradigm shift, from a more inclusive approach whether they are STEM or non-STEM, high or low academic level. 

Cody is a software coding platform tool.  Providing an intelligent and error-proof software coding learning system with easy-to-teach and easy-to-learn features for teachers and learners. This platform improves the Learning Experience (LE) of learners by applying a novel hybrid User-Interface that provide a graphical approach to text-based coding. The error-proof method helps to overcome the initial phase of coding skills development. Training teachers in the Target Market is another challenge. Our Cody platform encourages teachers’ adoption, by focusing on the Teaching Experience (TE) using curated teaching materials and assignment. Cody platform also incorporates Design Thinking tools that is integrated to supplement the LE and TE methods.

Introduction to Cody Features

1) Learning Management System

The platform is driven by the Learning Management System (LMS) serving both the Teachers, Students and Class Administrators roles.  Assignments between teachers and students are managed as a coding Projects. Classes, level of difficulty etc are managed by teachers using the LMS to post, review and mark the assignments. Teachers can also collaborate among themselves by sharing, editing, reusing the projects. A huge productivity gain for the teaching faculty.   

2) Virtual Coder Assistant

The coder assistant is virtualized into the Learner Experience, to ensure syntax are always correct. Learners will be exposed to the hybrid text-driven coding at the onset when using the Cody platform.  

Step (1) learners will select the type of software command they require in the Function Pane

Step (2) Virtual Coder Assistant Pane will immediately present the required parameters for learners to fill up.

Step (3) Once learners confirm the parameters entries in (2), the coder assistant will generate into the correct Python language and publish into the Text Editor Pane.

3) Teaching Tools

There are also various teaching tools build-into the platform. Teachers can create Assignments with partially hidden codes shown in (1) below. In the Student’s mode, student will exercise the computational concepts to derive the correct missing code. The same assignment template can also be modified with different hidden codes, and post to students as another assignment.  further enhancing the productivity of teachers in more variety of student assignments.   

4) Learning Progression

The teaching tools can provide a very rich learning progression for students by building up level of difficulty within assignments. Hidden codes can be gradually increased to build up students’ confidence in coding skills. Ultimately, students can develop codes with minimal guidance from teachers, achieve a good level of coding proficiency. 

5) Assessment

The platform can track in detail the speed and level of development in their coding skills. We observed that better learners can complete a correct assignment in a shorter time span. The tracking tool can be further developed using more sophisticated algorithm for automated assessment. The assessment can identify areas of strength or weakness of each learner. So that teachers can provide more suitable assignments accordingly. The assessment can also be known to the learners, for more self-directed learning.   

Cody Technology & Patent

Cody is developed based on in-house propriety technology. The design is primarily driven by various learning pedagogies and algorithmic programming methods. So that the system is scalable to support to a large number of users with a fast Virtual Coder Assistant. A patent was filed in Singapore 12th Oct 2021.

A conference abstract was also submitted and accepted by a leading university in Singapore, Singapore Institute of Technology’s (SIT) Applied Learning Conference (ALC) in 2022. (Loke, Bernard (2022): Ensemble of Teaching/Learning Pedagogies Applied to Software Coding to Enhance Learning Outcome. Singapore Institute of Technology. Conference contribution. https://doi-org.ezproxyberklee.flo.org/10.25447/sit.1... )

Cody Research & Collaboration

The Cody team attracted many interested collaborations in local university and polytechnic in Singapore. These institutions were excited by the Cody ability to improve the learning process and learning experience. In line with the fast pace development in Machine Learning and data analytics, R&D into software coding assessment attracted much attention among academia in the recent years. We are currently in discussion with 2 institutions of higher learning and 1 government agency in exploring how to adopt Cody in teaching coding to the youth together with a built-in assessment tool.    

The optimal age group for Cody is from K8 onwards. However, from our pilot trials, learners as young as K4 are capable of using Cody to learn basic coding skills by solving K4 math problems. One of the unique Cody features is applying math related problems and coding into an integral teaching/learning platform. Complementing computational thinking and coding skills, greatly lowering the learning barriers.

Current Situation in Learning Coding

The current spectrum of software learning tools for novice is typically heavily graphical driven at one extreme and a full-text driven for the experience coder at the other extreme. There is a huge void in the middle for novice learners. The graphical driven method, is usually catered for < K8 students. Older learners are beginning to use these graphical methods due to the lack of alternatives. However, the major drawback of graphical methods is the limitation when performing higher computational complexity coding task. Learners can face hurdle when crossing over to full-text driven coding subsequently. For the older learners, they tend to faze by the full-text coding at the onset resulting in high drop-out rate. 

Learners’ Profile

The learners are pre-dominantly segregated into 3 groups. First are the self-directed and talented learners. This group belongs to the 90th percentile and serve as the guru in the IT industry. The second group is the affluent learners, who can afford 1-to-1 teachers to learn coding. The third group is the non-affluent learners, who lacks the mental aptitude of the first group and the rich resources to support their learning aspiration. Majority of the youths are in the third group. For the supporting economies (as described earlier), the percentage of youths in the third group is even higher. Without any intervention, the disparity among the 3 groups would accelerate further as digital transformation progress around the world. This accelerative effect is already happening in developed countries.     

Cody Product Vision

Cody is designed to be a cost-effective program to learn coding, without extensive STEM education or high academic achievement. The core team and research collaborators all aligned towards a skill-based approach to learning coding instead of the traditional knowledge-based approach. We build the computation thinking upon K6 basic arithmetic skills, which is highly accessible to most learners in developing or developed countries. Applying our learning pedagogy together with the various Student/Teacher tools within Cody, as the next scaffold to build up the skills. Teaching examples and hands-on example will be adapted to daily lives’ context. Various level of difficulty can be design using more difficult context or more “hidden” codes.

Career Preparation Program – Cody Ascent Program

For learners completing the Cody program with proficient coding skills, can advance towards career development. This is an extension from program, Cody Ascent Program. This program is currently in development in Singapore. This extension program will introduce various software tools and best-practices used by the industry’s IT practitioners. Based on our in-house enterprise software development experience, we curate the best-of-breed most widely adopted tools. We will also modify our cloud-based software into a learning environment, whereby learners can apply skills and tools into this complete software environment for project-based curriculum. Essentially a 2-in-1 program, learning and on-job-training concurrently. Learners will be assigned a role in a development team and function as an integral team for each project assignment.  These industry software tools are increasing user-friendly but not learner-friendly. They are not learner-friendly because of the protocol dependent, making self-directed learning difficult especially those without career experience. Therefore, a skill-based oriented and project-based training is an optimal approach for effective show-and-tell. Following a structured program and sequencing the curated  industry tools leads into a much easier learning pathway. 

The company is a startup in Singapore focus in tech social-enterprise. For the past 5 years, the team members have been developing advance software system for the healthcare industry. The team comprises of software and algorithm experts. Many of our healthtech features and concepts were adopted by the mainstream companies. The system design is still the fastest in the market to register a hardware sensors to the cloud system (in seconds rather than in minutes) compared to competitors’ product. 

How the Journey Began

In 2020, the Team Leader (a non-programmer) was intrigued by the prowess of software development and the immense digital transformation momentum due to the COVID-19. Being a dyslexic himself, he fully appreciated the difficulty in learning software coding. He expounded the idea of an easy-to-learn software coding method for the youth regardless of their academic achievements. Surrounded by team members who are expert in software development, the feedback was very negative. The conventional wisdom was unanimous; “there is no special way to learn software”, “only the better ones can”, “everyone has to go through the difficult path”. Undeterred, the team leader proceeded to work with a few interns.

A survey was commissioned to find out youth’s learning motivation and susceptibility in learning software coding. More than 100 youth responded, and the results were surprising and overwhelmingly positive. Although most of the respondents were non-STEM students, the level of interest in learning coding was disproportionately higher. 

A trial class (titled: Cracking the Code: The link between Design Thinking and Coding) was organized and close to a 100 youth attended. All participants were surveyed after the trial class. For understanding and applicability of Design Thinking to software coding, the score was 4.28/5.0. Whereas understanding the introductory concepts to coding, the score was significantly lower at 3.5/5.0. There was limited software tool used during this class (Cody was not developed yet), probably contributed to the low score of 3.5/5.0. 

Cody Development Journey

Taking into considerations the feedback and comments from the survey and trial class, the Cody tool begin full development. Extensive research into various pedagogies were also applied to the developmental process.  Some of the design process concepts and principles can be found in this conference abstract that was submitted and accepted by a leading university in Singapore, Singapore Institute of Technology’s (SIT) Applied Learning Conference (ALC) in 2022. (Loke, Bernard (2022): Ensemble of Teaching/Learning Pedagogies Applied to Software Coding to Enhance Learning Outcome. Singapore Institute of Technology. Conference contribution. https://doi-org.ezproxyberklee.flo.org/10.25447/sit.1... )

The Cody software platform was subsequently used to teach close to 500 youths in Singapore. From the youngest in K4 to K12. The results were very positive. The ease of use of the Cody tool itself provided a good learning experience for learners to build up their confidence steadily. All the users of Cody were able to perform text-driven coding within a few hours of lesson. The wide age group also supported the concept of skills-based approach of Cody, instead of the traditional knowledge-based approach in teaching software coding. Please refer to the videos for users interview.

Uniqueness of Cody Team

Taking a contrarian attitude from the onset of Cody development, the Cody designed was driven emphatically to cater to the needs of coding novice in mind. Novel ideas without constraint was enjoyed by the initial team, notably from the team leader. Conversely, if the Cody development was undertaken by a team of software experts, the traditional biases and fundamental teaching approach may not be so easily succumbed.

The other success factors were in engaging the correct type of learners who are non-STEM oriented. The young learners, in K4 to K6, was also very helpful to emulate weaker academic students in higher education in other developing countries. Satisfying these latent groups of learners proved to be much more challenging than expected. However, the obstacles provided a clearer learning pathway for the team and better Cody features.

Currently, the entire developmental team is aligned to the Cody approach.

Communities & Ecosystem

The team places high importance in collaborating and partnering with educators in institute of higher learning in Singapore. Using our strong network from our R&D background, we are continuing to expand the network among tertiary institutions. We are beginning to engage the career centres tertiary institutions to start engaging youths prior to graduation. These efforts can further strengthen our links to the youth and to understand their needs. We are also currently working with a youth related government agency in engaging youth to prepare for their career especially in the Infocomm industry. The plan is to further expand and reach-out to STEM and non-STEM youth to enlarge digital skills enabled youth around the regions in Asia.

To further enrich the communities and ecosystem, the Cody program also plan to introduce learn-to-earn, whereby students can contribute by developing assignments to other students or if suitable to be an assistant teacher. Both the assignments or assistance teacher have financial compensation to encourage learners of downstream opportunities. 

The primary reason is to gain endorsement from this challenge. Solve has a strong brand recognition and network. We are keen to tap on the network and partners in Solve ecosystem. Especially in expanding beyond Asia, the network will be even more useful.

Secondary is the monetary support which will be helpful to offset the cost. 

Current software platform has limited success for non-STEM or non-tertiary educated youth. Most of the current software course cater to the needs of STEM and higher achiever youth who are unfazed with learning coding.

In contrast, Cody platform, targets to reach-out to youths who are non-STEM and lower academic achievers. Expanding to these areas can enhance their career opportunities in the digital economy. The software industry will have more talents to support the industry. The catalytic effects is huge, whereby non-STEM youth far out-number STEM youth. Another important impact will be more digital talents available to the industry.

Refer to the Demo video for users’ feedback. 

We aim to provide software coding competencies to youth in Singapore for Year 1. Also for Cody extension program (pre-career preparation course) starting in Year 1. From Year 2 onwards, we plan to expand overseas. Tentatively to 1 new country every year thereafter.

The Cody platform has a very strong analytics capability in measuring coding capabilities. Assignments are categorized into Easy, Medium and Difficult. The platform database is able to measure user performance. The performance measurements are based on time-to-completion of assignment, number of attempts etc. The learning curve or teaching method will benchmark different groups of learners against a calibrated database. Teaching methods can be modified accordingly based on the measured performance of each learners.

We are also working with a university partner in developing assessment methods. The assessment method will progress towards automation and AI. The vision is for that learners to enjoy a granular feedback on specific strength/weakness, without the cost-intensive manual grading or review methods.   

In essence, the measurable indicators after attending the course, will be provided by Coding Competency Benchmark (in terms of users’ Performance, Strength, Weakness),

Outcome:

To achieve digital literacy in software coding for STEM, non-STEM, High or Low academic level youths.

Supporting Evidence:

We have extensive evidence during the research and trial. With close to 500 students from K4 to K12 acquiring ability to use the software platform to perform software coding. We are able to demonstrate, the effectiveness of skill-based driven approach of Cody, vs the traditional knowledge-based approach used in traditional software learning methodologies.

Refer to conference abstract that was submitted and accepted by a leading university in Singapore, Singapore Institute of Technology’s (SIT) Applied Learning Conference (ALC) in 2022. (Loke, Bernard (2022): Ensemble of Teaching/Learning Pedagogies Applied to Software Coding to Enhance Learning Outcome. Singapore Institute of Technology. Conference contribution. https://doi-org.ezproxyberklee.flo.org/10.25447/sit.1... )

Core Technology

User-Interface with a menu of selectable software functions, with parameters input and instantaneous transformation to proper Python text-code. The core technology resides in the automated transformation from structure information into Python correct syntax code.

Scalability

The applicability of the software platform is browser-based for laptop users or in smart-app for mobile users. The user-interface is adaptable for both laptop and smartphone format. The user is also not limited to any number since the system infrastructure is cloud-based.

AI Development

Further R&D into applying AI into the Cody platform for more features.

Intellectual Property

Patent Title: “SOFTWARE PLATFORM FOR SOFTWARE CODING WITH LEARNING

MANAGEMENT SYSTEM FOR TEACHERS AND LEARNERS USING

INTELLIGENT CODE GENERATION METHOD”, filed in Singapore 12th Oct 2021.

Refer to the conference abstract and the demo video.

Conference abstract that was submitted and accepted by a leading university in Singapore, Singapore Institute of Technology’s (SIT) Applied Learning Conference (ALC) in 2022. (Loke, Bernard (2022): Ensemble of Teaching/Learning Pedagogies Applied to Software Coding to Enhance Learning Outcome. Singapore Institute of Technology. Conference contribution. https://doi-org.ezproxyberklee.flo.org/10.25447/sit.1... )

The technical team members are mostly engineering trained and based wholly on merit and capabilities. We value diversity and inclusiveness. Not all members are software trained, providing different engineering approach. We also have a member who was previously physically handicapped for many years and recently recovered. We also rely heavily on new graduates as intern, to value their new engineering perspective and their youthful experiences.  Some of our members are also located overseas. 

Majority of the development is done remotely due to the COVID-19 situation. The leadership was acutely aware of the challenges in communication and collaboration at the onset. To breakdown potential barriers, various communication and collaboration tools were used, like Zoom, Google tools, Slack etc.  Work-flow is also designed specifically to ensure optimal productivity. The experience was better than we expected. Especially in terms of inclusiveness, where we can maintain high level of respect and listening skills without compromising on the outcome.

In the next phase of team expansion, we plan to recruit youths from non-STEM and medium to low academic achievement to join our team. Depending on the overseas expansion plan, we will also employ locals within the market using similar criteria.

Another target market we are preparing is the seniors (>50 years old) who are software illiterate. We recently held a 2 hours trial lesson. Some of the participants are also to code an Medium level assignment using for K4 to K6 materials.

In addition, we also incorporate an adult trainer (> 50 years old), who is software illiterate. He is currently undergoing training on software concepts. The plan is for teach software concept to the adult seniors learners. 

The revenue will be a combination of fee-based students, contribution-in-kind students, and corporate sponsorship.

For fee-based students, there will be no requirement for the students to perform contribution in-kind services. For non-paying students, a contribution-in-kind services will be expected by performing teaching assignment or developing course materials. Corporate sponsors can also enjoy talent recruitment from the graduate pool of students from the program.

The product will be conducted virtually. Our Cody platform is able to offer the learning convenience, especially for remote locations in some countries. For the non-paying students, to further reduce cost, they can also use smartphone to participate in the lessons as well as doing assignment using Cody smart-app.

For promotion, the Cody platform allows learners to email their assignments to their friends to share their achievements. The friends are also encouraged to sign-in to the Cody platform and on-board to the class subsequently. Creating a community led ecosystem. The promotion strategy will leverage on the 2 schemes for students: fee-based and contribution-in-kind.

The advantage of the business model is the high demand for digital literacy. Especially for the non-knowledge-based youth with limited academic capabilities. Leveraging on various social channels to reach out to the less affluent youth population, we can conveniently provide online streaming to market our product. Exhibiting the value proposition to the fee and non-fee-based students. Followed by online enrolment through browser or smart-app. Creating a frictionless process from marketing, onboarding to class delivery.

We are currently marketing regular fee-based class from Q2 2022 onwards. Based on the fee, would be positive operating profit margin, after covering trainers’ cost and marketing. Courses will also be operating with new government grant, riding on previous successful grant application.

On a longer term, we plan to attract external investors for fast growth phase. In fast growth phase, the plan is to increasing marketing in Singapore and overseas as well as technology build-up for the Cody platform.

We bootstrap throughout the 2021 developmental phase of the Cody platform. In Q4 2021 also received limited revenue from trial class. 

In Q1 2022, we have successes in various initiatives. We have successfully received grant approval from government agencies. Our courses have also received approval for Skillsfuture in Singapore with stringent criteria to provide high learning value courses (Skillsfuture is a government scheme in Singapore, whereby for approval courses, eligible participants can enjoy fees credited from this scheme, instead of paying the fee by the participants).

We have also garnered many interested partners from government related tertiary educators to collaborate on technical development. This is a strong validation on the potential success of the Cody platform.

Afternote: most Education Technology (EdTech) in the Asia region does not achieve the level of sophistication to attract tertiary educators. Most EdTech products are usually service-oriented instead of technology-oriented. Another group of EdTech products are curriculum-oriented developed by practitioners in the K1 to K12 teaching faculty and IT experts providing IT bootcamps.