" Select an industry role, assume yourself to be part of PreProd Corp's development team to design, develop, test, deploy and maintain actual AI pipelines for industry applications using the Emulet Bundles "
As an Emulet learner, you will engage in all phases of project work, including procurement, design, development, testing, deployment, and maintenance of real-world AI pipelines as practiced in the industry. These phases are packaged as different bundles under Emulet Program.
STEP 01
Introducing Client - Vendor
A formal introduction is made between two entities: the first, referred to as the "Client," is a company seeking technology development services to support and enhance its business operations. The second, designated as "PreProd Develop" for all subsequent references, is the organization providing the required technology development services.
STEP 02
Business Problem
A comprehensive overview of the Client's current business challenge is provided, including an analysis of the root cause, the impact on business operations, and the proposed solution formulated by the Client's product technology team. Additionally, the preliminary timeline and budget approved by the Client's C-suite executives are outlined. The discussion emphasizes the urgency expressed by the Client's product technology team to engage an external technology development company to design and implement an AI-driven solution addressing the identified challenge.
STEP 03
Capability Presentation
PreProd Develop prepares a tailored presentation to showcase its technological expertise and delivery capabilities, aimed at enabling the Client's C-suite executives to make informed decisions regarding advancing discussions. Furthermore, PreProd Develop conceptualizes, designs, and demonstrates a custom Minimum Viable Product (MVP) closely aligned with the Client’s envisioned solution to address their current business challenge. This demonstration provides the Client with a tangible experience of the proposed AI-driven solution while highlighting PreProd Develop's agility and technological proficiency, reinforcing the value of a potential partnership.
STEP 04
RFQ & RFP Processes
Following prior interactions, the Client extends an invitation to PreProd Develop to submit a Request for Quote (RFQ), seeking an approximate cost estimate for developing the required AI solution. PreProd Develop is responsible for preparing and submitting the necessary documentation within the specified timeline.
Subsequently, based on the RFQ submission, the Client invites PreProd Develop to provide a comprehensive Request for Proposal (RFP). PreProd Develop must ensure the timely completion and submission of all requisite documentation.
These processes are highly confidential and critical to both organizations.
STEP 05
Signing a SOW
PreProd Develop successfully secures a business contract with the Client to develop an AI solution addressing their current business challenge. Authorized representatives from both organizations formally sign a Statement of Work (SOW), which comprehensively outlines and details all critical information pertaining to the engagement.
BUNDLE 01
PROCUREMENT
This bundle provides hands-on working experience into the process of finalizing a Statement of Work (SOW) contract between clients and service providers. Gain a practical understanding of securing business for a real AI project by immersing yourself (emulate) in the role of PreProd Develop.
STEP 01
Business Analysis
PreProd Develop engages an experienced Business Analyst tasked with collaborating closely with the Client's team. The Analyst's responsibilities include identifying and engaging relevant stakeholders, documenting the current versus desired state, developing a functional understanding of available data, and gathering all pertinent information to ensure alignment between PreProd Develop and the Client's objectives.
STEP 02
Design a POC
PreProd Develop efficiently designs and develops a high-level working model using sample data closely resembling the Client's data. PreProd Develop delivers a comprehensive demonstration of the Proof of Concept (POC) to the Client, highlighting their technical capabilities, validating the Client's concept, and meticulously documenting all encountered challenges for future retrospection.
STEP 03
Design Architecture
PreProd Develop collaborates with the Client's technology team to design a comprehensive system architecture. This architecture encompasses all critical components, including data sources, ingestion, enrichment, processing, storage, dashboards, reporting, explainable AI (XAI) model training, MLOps, and deployment pipelines. The design ensures seamless integration of tools and systems, scalability for future upgrades or replacements, and uninterrupted operations to effectively address the Client's business challenge.
PreProd Develop navigates challenges and blockers during the design process, carefully selecting and integrating components while considering factors such as budget constraints, required skills, timelines, compatibility with the existing tech stack, and provisions for future scalability and upgrades.
PreProd Develop establishes a baseline of multiple architectural designs, detailing the components, data flow, triggers, code flow, and integrations, ensuring the solution aligns with the Client's business objectives and resolves the identified problem.
STEP 04
Delivery Plan
PreProd Develop, in collaboration with the Client's delivery team, creates a comprehensive delivery plan that defines the project scope, outlines the onboarding of the appropriate team, sets key milestones, and establishes processes for release management, budget oversight, timesheet tracking, monitoring, reporting, communication, and all other standard operating procedures (SOPs) necessary for the successful execution of the project.
Additionally, PreProd Develop ensures that all internal stakeholders involved in the delivery are well-equipped with a clear understanding of the project scope, Client requirements, expected solution, timeline, product roadmap, agile methodologies, project documentation, time logs, communication protocols, conflict resolution strategies, and all other best practices recommended by both the internal team and the Client to guarantee the project’s success.
STEP 05
Product Board Setup
PreProd Develop, in collaboration with the Client's technology team, conducts a thorough requirements engineering process by categorizing high-level, mid-level, and detailed requirements, while creating an accurate traceability matrix to facilitate effective change management. PreProd Develop breaks down these requirements into smaller user stories and larger epics, ensuring readiness and flexibility for the development, testing, and ongoing maintenance of the entire project.
Additionally, PreProd Develop meticulously documents all user stories on an agile board, specifying their priority, story points, and sprint plan—detailing which user stories will be developed in each sprint.
BUNDLE 02
Initiate projecT
This bundle provides hands-on working experience into the process of initiating an AI development project. Gain a practical understanding of setting up all required plans and processes for a real AI project by immersing yourself (emulate) in the role of PreProd Develop.
STEP 01
Spike Activity
PreProd Develop performs a spike event to understand the client's current tech stack, all available data sources, data mapping, data functionality and any other details that will help the team to thoroughly understand the existing data. Identify the delta between the system design architecture vs the current skill matrix of the team and fill the delta by training the team on the required skills. The team performs R&D on the tools and technologies that is going to be used prior to starting with the actual development.
STEP 02
Infrastructure Setup
PreProd Develop performs a complete infrastructure setup for data engineering pipeline development. This is a highly challenging task as the team has to carefully choose the operating system, databases, computational environments and compatible programming or scripting languages, libraries and all other required project environments within their existing setup, enabling them to initiate and deliver the project according to the client's expectations.
STEP 03
Pipeline Development
PreProd Develop develops a robust data engineering pipeline consisting of configuring data sources, data streaming, data ingestion, data transformation (basic), data enrichment (functional), data processing and efficient storage making the enriched data available for all data consumers.
STEP 04
Pipeline Testing
PreProd Develop performs functional, performance, security and integration testing for the data engineering pipeline on volume, velocity, variety and veracity of the transformed data. Testing team shares all feedback and identified bugs and recommended transformation logic with the development team to get them fixed.
STEP 05
Pipeline Demo
PreProd Develop invites all the internal and external (client) stakeholders for demonstrating the developed data engineering pipeline. PreProd Develop collects all the possible feedbacks from the stakeholder ensuring they are implemented on the developed pipeline (if feasible).
BUNDLE 03
DATA ENGINEERING
This bundle provides hands-on working experience into the process of designing, developing and testing a data engineering pipeline. Gain a practical understanding of setting a robust data engineering pipeline for a real AI project by immersing yourself (emulate) in the role of PreProd Develop.
STEP 01
Spike Activity
PreProd Develop performs a spike event where the team members spend some time to understand the client's current tech stack when it comes to designing AIML models. Understands all available data sources, data mapping, data functionality and any other details that will help the team to thoroughly understand the existing data and all possible AIML models that can be trained on the data. Identify the delta between the system design architecture vs the current skill matrix of the team and fill the delta by training the team on the required skills. The team performs R&D on the tools and technologies that is going to be used prior to starting with the actual development.
STEP 02
Infrastructure Setup
PreProd Develop performs a complete infrastructure setup for AIML training development environment. This is a highly challenging task as the team has to carefully choose the operating system, databases, computational environments and compatible programming or scripting languages, libraries and all other required project environments within their existing setup, enabling them to initiate and deliver the project according to the client's expectations.
STEP 03
Pipeline Development
PreProd Develop designs and develops a XAI pipeline. Experience procuring the transformed data from data engineering team, data transformation and new feature generation out of current data which is compatible with AIML model/s that are currently being experimented by performing exhaustive training and tuning for getting the best possible results. Development of production ready AIML model/s and adds a layer of explainability over it ensuring it provides most accurate results as predictions.
STEP 04
Pipeline Testing
PreProd Develop perform testing and validation for the developed XAI models. Experience model validation on the developed XAI model/s using the controlled live data from the client i.e. client already has provided the results which will be compared with the predictions provided by the XAI model/s. Testing team shares the all feedback and identified delta in the performance with the development team to get them fixed.
STEP 05
Pipeline Demo
PreProd Develop invites all the internal and external (client) stakeholders for demonstrating the developed XAI pipeline. PreProd Develop collects all the possible feedbacks from the stakeholder ensuring they are implemented on the developed pipeline (if feasible).
BUNDLE 04
EXPLAINABLE AI
This bundle provides hands-on working experience into the process of designing, developing and testing a AIML model training pipeline. Gain a practical understanding of setting a robust explainable AI (XAI) pipeline for a real AI project by immersing yourself (emulate) in the role of PreProd Develop.
STEP 01
Spike Activity
PreProd Develop performing a spike event where the team members spend some time to understand the client's current tech stack when it comes to designing machine learning operations pipeline. Understands all the tools, technologies and processes employed by the data engineering and XAI pipelines. Identify the delta between the system design architecture vs the current skill matrix of the team and fill the delta by training the team on the required skills. The team performs R&D on the tools and technologies that is going to be used prior to starting with the actual development.
STEP 02
Infrastructure Setup
PreProd Develop performs a complete infrastructure setup for setting up a machine learning operations environment. This is a highly challenging task as the team has to carefully choose the operating system, databases, computational environments and compatible programming or scripting languages, libraries and all other required project environments within their existing setup, enabling them to initiate and deliver the project according to the client's expectations.
STEP 03
Pipeline Development
PreProd Develop performing development for the MLOps pipeline. Experience developing schedulers for running data engineering pipeline, drift monitoring system, trigger for XAI model training, XAI model saving and serving the same to the end API. Experience implementing production ready MLOps pipeline ensuring it streams data, processes data, performs AIML model training and provide with most accurate results as predictions.
STEP 04
Pipeline Testing
PreProd Develop performing integration testing between the developed data engineering, XAI and MLOps pipelines. Experience how the technology team streams live data from the client's application data sources into data engineering pipeline and from there to XAI pipelines to train, test, validate and store the all/best trained models for consumption all these triggered and managed by MLOps pipeline.
STEP 05
Pipeline Demo
PreProd Develop invites all the internal and external (client) stakeholders for demonstrating the developed MLOps pipeline. PreProd Develop collects all the possible feedbacks from the stakeholder ensuring they are implemented on the developed pipeline (if feasible).
BUNDLE 05
ML Operations
This bundle provides hands-on working experience into the process of designing, developing and testing a MLOps pipeline. Gain a practical understanding of setting a robust operations pipeline for a real AI project by immersing yourself (emulate) in the role of PreProd Develop.
STEP 01
Spike Activity
PreProd Develop performing a spike event where the team members spend some time to understand the client's current tech stack required to match client's Pre-production environment. Understands all the tools, technologies and processes employed by the client. Identify the delta between the system design architecture vs the current skill matrix of the team and fill the delta by training the team on the required skills. The team performs R&D on the tools and technologies that is going to be used prior to starting with the actual development.
STEP 02
Integrate Data Eng. & XAI pipeline
PreProd Develop performing integration between the developed XAI pipeline with data engineering pipeline. Experience how the technology team streams live data from the client's application data sources into data engineering pipeline and from there to XAI pipelines to train, test, validate and store the training pickle for future use. Experience the demo to the client by showcasing the data engineering and XAI pipelines working together to produce predictions on client's data. All the feedback and recommendations shared by the client for enhancing the performance of the pipelines and predictions is shared with the team to make the recommended fixes.
STEP 03
Integrate MLOps
PreProd Develop performing a deployment for implementing MLOps pipeline on pre-production environment. Experience the team works closely with the client's application team to gain all the knowledge and access to client's pre-production environment. Experience deploying and monitoring production ready MLOps pipeline onto client's pre-production environment ensuring it accepts live requests for predictions from the client's application, triggers the serving XAI model to output predictions. In parallel MLOps pipeline also streams live data and makes a decision to retrain the XAI model if required ensuring the client application always interacts with latest XAI model. Experience the demo highlighting the results and reports by running the MLOps pipeline on the pre-production environment. Once the client's technology team gives "Go Ahead", only then the MLOps pipeline is deployed on client's actual production environment.
STEP 04
Monitoring & Reporting
PreProd Develop performing a deployment for implementing monitoring and reporting pipeline on pre-production environment. Experience the team works closely with the client's application team to validate the dashboards and reports. Once the client's technology team gives "Go Ahead", only then the MLOps pipeline is deployed on client's actual production environment.
STEP 05
Pipeline Demo
PreProd Develop invites all the internal and external (client) stakeholders for demonstrating the performance of the deployed pipelines. PreProd Develop collects all the possible feedbacks from the stakeholder ensuring they are implemented on the developed pipeline (if feasible).
BUNDLE 06
DEPLOYMENT
This bundle provides hands-on working experience into the process of integrating data engineering, XAI, MLOps and BI pipelines and later deploying the integrated pipelines onto client's pre-production environment. Gain a practical understanding of setting pipeline deployment and monitoring onto preproduction environment for a real AI project by immersing yourself (emulate) in the role of PreProd Develop.
STEP 01
Spike Activity
PreProd Develop performs a spike event where the team members spend some time to understand the current tech stack i.e. the data engineering, XAI and MLOps pipelines deployed on client's production environment when it comes to designing monitoring and reporting pipeline. Understands all the tools, technologies and processes employed by the data engineering and XAI pipelines. Identify the delta between the existing architecture vs the current skill matrix of the team and fill the delta by training the team on the required skills. The team performs R&D on the tools and technologies that is going to be used prior to starting with the actual development.
STEP 02
Infrastructure Setup
Experience PreProd Develop performing a complete infrastructure setup for monitoring and reporting development environment. This is a highly challenging task as the team has to carefully choose the operating system, databases, computational environments and compatible programming or scripting languages, libraries and all other required project environments within their existing setup, enabling them to initiate and deliver the project according to the client's expectations.
STEP 03
Pipeline Development
PreProd Develop performing development for the monitoring and reporting pipeline. Experience developing live interactive dashboards for monitoring data engineering pipeline, data drift, model training, latest and historical model performance measures, computational performance, etc. Experience developing customized automated reports for data engineering pipeline, XAI and MLOps pipelines as per client's dynamic requirements.
STEP 04
Preprod Deployment
PreProd Develop performing a deployment for implementing monitoring and reporting pipeline on pre-production environment. Experience the team works closely with the client's application team to validate the dashboards and reports. Once the client's technology team gives "Go Ahead", only then the MLOps pipeline is deployed on client's actual production environment.
STEP 05
Pipeline Demo
PreProd Develop invites all the internal and external (client) stakeholders for demonstrating the deployed BI pipeline. PreProd Develop collects all the possible feedbacks from the stakeholder ensuring they are implemented on the developed pipeline (if feasible).
BUNDLE 07
Business Intelligence
This bundle provides hands-on working experience into the process of designing, developing and testing a monitoring and reporting BI pipeline. Gain a practical understanding of setting a robust BI pipeline for a real AI project by immersing yourself (emulate) in the role of PreProd Develop.
STEP 01
Priority Four Incident
One of the deployed pipelines in client's production environment is not functioning as expected and may result in a minor impact to client's business in near future. Experience how PreProd Develop is assigned with a task to find the root cause, fix the issue and document all the relevant information as per the SLA agreed in the signed SOW.
STEP 02
Priority Three Incident
One of the deployed pipelines in client's production environment is not functioning as expected and will result in a considerable impact to client's business in near future. Experience how PreProd Develop is assigned with a task to find the root cause, fix the issue and document all the relevant information as per the SLA agreed in the signed SOW.
STEP 03
Priority Two Incident
One of the deployed pipelines in client's production environment is not functioning as expected and will result in a serious impact to client's business on immediate basis and may result in temporary downtime. Experience how PreProd Develop is assigned with a task to find the root cause, fix the issue and document all the relevant information as per the SLA agreed in the signed SOW.
STEP 04
Priority One Incident
One of the deployed pipelines in client's production environment is not functioning as expected and will result in a serious impact to client's business on immediate basis and has resulted in downtime. Experience how PreProd Develop is assigned with a task to find the root cause, fix the issue and document all the relevant information as per the SLA agreed in the signed SOW.
STEP 05
Incident Reporting
PreProd Develop creates, manages, and updates the necessary documentation, detailing the root cause analysis (RCA), the solution implemented to address the issue, best practices to prevent recurrence, turnaround time, SLA compliance, and other relevant metrics. This documentation serves to inform and align all stakeholders effectively.
BUNDLE 08
INCIDENT MANAGEMENT
This bundle provides hands-on working experience into the process of dealing with different levels of production related incidents. Gain a practical understanding of identifying and providing quick and robust solutions ensuring minimal product downtime for a real AI project by immersing yourself (emulate) in the role of PreProd Develop.
STEP 01
Data Values
PreProd Develop receives one or many change requests (CR) as a result of a critical application data being modified (added or deleted) on the business side. After careful analysis, PreProd Develop provides a timeline to client for making required changes first on development environment and later it is handed over the testing team for testing. The modified pipeline is deployed on pre-production environment under observations. Once it qualifies all the criteria, the results are demonstrated to the client and upon their approval the same is pulled on client's production environment.
STEP 02
Business Logic
PreProd Develop receives one or many change requests (CR) as a result of a critical business logic being modified on the business side. After careful analysis, PreProd Develop provides a timeline to client for making required changes first on development environment and later it is handed over the testing team for testing. The modified pipeline is deployed on pre-production environment under observations. Once it qualifies all the criteria, the results are demonstrated to the client and upon their approval the same is pulled on client's production environment.
STEP 03
Project Infrastructure
Experience how PreProd Develop receives one or many change requests (CR) as a result of a critical infrastructure being modified on the technology operations side. After careful analysis, PreProd Develop provides a timeline to client for making required changes first on development environment and later it is handed over the testing team for testing. The modified pipeline is deployed on pre-production environment under observations. Once it qualifies all the criteria, the results are demonstrated to the client and upon their approval the same is pulled on client's production environment.
STEP 04
Feature Enhancement
Experience how PreProd Develop receives one or many change requests (CR) as a result of strategy for generating new functional features from the existing available data being implemented on the data operations side. After careful analysis, PreProd Develop provides a timeline to client for making required changes first on development environment and later it is handed over the testing team for testing. The modified pipeline is deployed on pre-production environment under observations. Once it qualifies all the criteria, the results are demonstrated to the client and upon their approval the same is pulled on client's production environment.
STEP 05
External Tool Integrations
Experience how PreProd Develop receives one or many change requests (CR) as a result of a critical external tool being integrated on the product side. After careful analysis, PreProd Develop provides a timeline to client for making required changes first on development environment and later it is handed over the testing team for testing. The modified pipeline is deployed on pre-production environment under observations. Once it qualifies all the criteria, the results are demonstrated to the client and upon their approval the same is pulled on client's production environment.
BUNDLE 09
CHANGE REQUESTS
This bundle provides hands-on working experience into the process of dealing with variety of critical change requests resulting due to change in the vision, business and infrastructure from the client side. Gain a practical understanding of identifying and providing quick and robust solutions ensuring incorporation of all necessary changes to the existing pipelines having minimal product downtime for a real AI project by immersing yourself (emulate) in the role of PreProd Develop.
STEP 01
Data Engineering
Experience how PreProd Develop proactively designs an automation POC (not using any of the the client's resources) to automate the entire data engineering process by creating a minimal GUI that can replace entire manual trigger and execution to operate the data engineering pipeline and showcases a considerable ROI in future for the client if implemented. Upon client's approval and allocation of additional budget, the automation is developed, deployed on client's pre-production environment and later onto client's production environment.
STEP 02
AI Model Training
Experience how PreProd Develop proactively designs an automation POC (not using any of the the client's resources) to automate the entire XAI pipeline process by creating an AutoML GUI that can replace entire manual trigger and execution to operate the AIML model training pipeline and showcases a considerable ROI in future for the client if implemented. Upon client's approval and allocation of additional budget, the automation is developed, deployed on client's pre-production environment and later onto client's production environment.
STEP 03
Data Drift Monitoring
Experience how PreProd Develop proactively designs an automation POC (not using any of the the client's resources) to automate the entire data drift monitoring process by creating a minimal GUI that can replace entire manual trigger and execution to operate trigger for AIML model training and showcases a considerable ROI in future for the client if implemented. Upon client's approval and allocation of additional budget, the automation is developed, deployed on client's pre-production environment and later onto client's production environment.
STEP 04
API Development + Integrate
Experience how PreProd Develop proactively designs an automation POC (not using any of the client's resources) to provide a simple GUI that has AIML model serving on its backend. This GUI enables the client to manually enter the parameters using which the GUI will provide a near to accurate future prediction which could be used internally by client's team for various purposes. Upon client's approval and allocation of additional budget, the automation is developed, deployed on client's pre-production environment and later onto client's production environment.
BUNDLE 10
AUTOMATION
This bundle provides hands-on working experience into the process of designing and developing major/minor automations within the developed AI pipelines. Gain a practical understanding of identifying automation as a potential for a real AI project by immersing yourself (emulate) in the role of PreProd Develop.
