| Summary | Description | Component | Priority | Completion |
|---|---|---|---|---|
| AI Scaling | *Start Epic from Jan 1, 2025. Sprint 35-43 | Epic Due Date: May 31, 2025* *Key Result 7:* Bring/expand AI pilots to 10 more country contexts (5 low-resource languages covered), while ensuring AI tools are adapted to local needs and tested for reliability and scalability. This Epic focuses on scaling AI solutions to new country contexts, ensuring AI tools are effective in diverse environments. It includes prioritizing expansion opportunities, training local teams, adapting tools for low-resource languages, and monitoring progress to develop a robust scaling strategy. *Story Points:* Maximum Points across Developers and Non-Dev Contributors (30–40) across 9 Sprints - 4.5 months. *Deliverables:* # *Context Prioritization – End Date: February 28, 2025* * Identify and prioritize new country contexts for AI pilot expansion based on feasibility and impact assessments. * Engage stakeholders to understand local needs and tailor AI applications to specific humanitarian challenges. * Create a roadmap for scaling, detailing timelines, resources, and expected outcomes. # *Localization and Training – End Date: April 30, 2025* * Recruit and onboard local teams to implement pilots in selected country contexts. * Design and deliver training programs to equip local teams with the skills to manage AI tools effectively. * Adapt AI tools to support 5 low-resource languages, ensuring linguistic and cultural relevance. # *Pilot Implementation and Validation – End Date: May 31, 2025* * Deploy AI tools in new pilot contexts, focusing on low-resource environments. * Test and validate AI capabilities in each context to ensure reliability and usability. * Monitor and report on the progress of all new pilots, highlighting successes and challenges. # *Scaling Strategy Development – End Date: May 31, 2025* * Develop comprehensive scaling guidelines based on results from the new pilots. * Document lessons learned, best practices, and recommendations for future expansions. * Create a final report summarizing pilot outcomes and the roadmap for further scaling efforts. *Inclusions Checklist:* * Identify and prioritize new country contexts for AI pilot expansion. ✅ * Recruit and train local teams for pilot implementation. ✅ * Adapt AI tools to support low-resource languages. ✅ * Test and validate AI capabilities in new contexts. ✅ * Develop scaling guidelines based on pilot results. ✅ * Monitor and report progress across all new pilots. ✅ *Milestones and Deadlines:* * *Context Prioritization Complete:* February 28, 2025 * *Localization and Training Complete:* April 30, 2025 * *Pilot Implementation and Validation Complete:* May 31, 2025 * *Scaling Strategy Development Complete:* May 31, 2025 | AI | High | 16.67% |
| AI Research | *Start Epic from Jan 1, 2025. Sprint 35-44 | Epic Due Date: May 31, 2025* *Epic Due Date: May 31, 2025* *Key Result 4:* Establish signpostAI as the leading authority on applied use of generative AI research for humanitarian aid, publishing groundbreaking research, participating in thought leadership events, and fostering collaborations with key research institutions. X amount of ind. published research + X amount of collab pieces. This Epic focuses on advancing the reputation of SignpostAI as a thought leader in the humanitarian AI sector. It involves developing actionable guidelines, fostering collaborations, and publishing research outputs to shape the discourse on responsible AI for humanitarian applications. *Story Points:* Maximum Points across Developers and Non-Dev Contributors (30–40) across 10 Sprints - 5 months. *Deliverables:* # *Guideline Development – End Date: February 28, 2025* * Develop a humanitarian Responsible AI (RAI) guideline tailored for the aid sector (UKHIA?). * Conduct stakeholder interviews to align the guidelines with practical use cases and sector needs. * Publish the finalized guideline on the SignpostAI platform and promote it through blogs and newsletters. # *Research Collaboration – End Date: March 31, 2025* * Partner with key institutions to co-author collaborative research on generative AI applications. * Identify and secure contributors for joint research projects and thought leadership events. * Establish a recurring research schedule and deliver preliminary results for feedback. # *Publications and Outreach – End Date: May 31, 2025* * Publish at least three research pieces, including individual and collaborative efforts. * Organize and participate in at least two thought leadership events or conferences. * Promote research outputs via blog posts, social media, and newsletters. * Track and report research impact metrics, such as citations and mentions in humanitarian AI discussions. *Milestones and Deadlines:* * *Guideline Development Complete:* February 28, 2025 * *Research Collaboration Initiated:* March 31, 2025 * *Publications and Outreach Complete:* May 31, 2025 | AI | High | 27.78% |
| AI AutoEval | *Start Epic from Jan 1, 2025. Sprint 35-39 | Epic Due Date: May 31, 2025* *Key Result 3*: Create an automatic evalutation that can run on a % of inquiries so it doesn't require human eval 100% of the time. Aiming to decrease human time spent on evaluation by 80%. **How will we calculate? >>Develop deep learning ticket analysis based on the ticket validation framework already in place. This Epic focuses on designing and deploying an automated evaluation tool to streamline ticket validation processes. The goal is to reduce manual effort while maintaining the accuracy and efficiency of AI systems in real-world scenarios. *Story Points:* Maximum Points across Developers and Non-Dev Contributors (30–40) across 5 Sprints - 2.5 months. *Deliverables:* # *Framework Analysis –* End Date: *February 28, 2025* * Analyze the existing ticket validation framework to identify process gaps. * Develop a harm taxonomy to classify and mitigate potential risks associated with automated evaluations. * Define success metrics for evaluating the performance of the automated tool. # *Model Development –* End Date: *March 31, 2025* * Design a deep learning model to evaluate ticket inquiries based on the established validation framework. * Develop an automated evaluation tool to process a subset of inquiries. * Integrate the tool with Zendesk to align with existing workflows and user needs. # *Testing and Validation –* End Date: *May 31, 2025* * Test the tool on a controlled percentage of real-world inquiries to measure its effectiveness. * Validate the reduction in human effort using predefined success metrics. * Refine the tool based on feedback and testing results. *Milestones and Deadlines:* * *Framework Analysis Complete:* February 28, 2025 * *Model Development Complete:* March 31, 2025 * *Tool Testing and Validation Complete:* May 31, 2025 | AI | High | 15.38% |
| AI Confidence | *Start Epic from Jan 1, 2025. Sprint 35-40 | Epic Due Date: May 31, 2025* Key Result 1: Develop a comprehensive risk assessment, roadmap, and data evaluation framework to enable Quality decision-makers to recommend whether or not to use direct-to-client AI interactions all while ensuring Signpost quality and safety standards. This Epic focuses on creating and validating a risk assessment framework for direct-to-client AI interactions, ensuring robust quality and safety standards. It includes developing guidelines, models, and tools to support decision-making for deploying AI in humanitarian contexts while maintaining adaptability across diverse pilot locations. *Story Points*: Maximum Points across Developers and Non-Dev Contributors (30–40) across 10 Sprints - 5 months. *Deliverables:* # *Flow Design and Testing*: * Define and test flows to evaluate AI interactions for direct-to-client use cases. * Develop modular components that support both human-in-the-loop and automated processes. * Ensure flows are adaptable across pilot locations and humanitarian contexts. # *Risk Assessment Framework and Tools*: * Create a harm taxonomy to identify, categorize, and mitigate risks. * Develop a comprehensive risk assessment model. * Build a risk assessment bot integrated with Zendesk or agent handovers. # *Quality Assurance and Guidelines*: * Validate the quality of the risk assessment bot through rigorous testing. * Develop humanitarian Responsible AI (RAI) guidelines tailored to the aid sector. # *Data Analysis and Standards*: * Review pilot data and establish trends in content failures to set clear standards for when AI should respond vs. defer to human moderators. # *Go/No-Go Framework*: Select from limited testing options for go/no-go. * Design and test a go/no-go decision-making framework to evaluate the readiness of direct-to-client AI deployments. # *Pilot Deployments*: * Launch low-resource voice assistants and tools in existing pilot locations (e.g., Greece, Kenya, El Salvador). * Collect feedback and iterate based on pilot performance. *Milestones and Deadlines:* * *Flow Design and Testing Complete*: January 2025 * *Risk Assessment Bot Deployed*: February 2025 * *Go/No-Go Framework Finalized*: March 2025 * *Final Report on AI Deployment Decisions*: April 2025 | AI | High | 31.58% |
| Data Protection | Develop baseline standards for Data Protection | Medium | 0.00% | |
| Geospatial Targeting | Operationalize 3-4 use cases using geospatial data to automate decisionmaking and actions for direct service delivery with other teams (ERD, Climate, Health) _Operationalize 3-4 use cases alongside climate, cash, health_Connect Google Flood Hub API and other global climate warning systems with derisking protections _Connect Signpost’s Global Mapping hub main data layer for environmental, economic, and conflict risk data _Conduct test advertising campaigns to validate at-scale model with human agents paired with unsupervised AI " | Medium | 50.00% | |
| Digital Cash | Prove concept that we can send cash digitally to 100 people in a limited pilot context _Finalize contract Paperwork with Mastercard _Explore Stripe _Create one demonstration penny test with FX, transactions, and payments | Medium | N/A | |
| Data Classification | Create the evaluation and deep analysis of tickets, knowledge base, and other factors that allow us to determine our confifence in AI direct to client _Run analysis of LLM on entire knowldege base and historical messages to create a classification schema _Create revised classification schema with definitions _Create a trends analysis of historical tickets _Create an automated trends analysis format in global dashboard _Create dashboard ability to analyze error and quality of messages based on knowledge base and AI variables and ZD ticket factors _Create a visualization and tooltip recommendations detailing success rates and gaps in knowledge base _ Test autotagging in Zendesk in pilot contexts _roll out autotagging of all tickets | Medium | 0.00% | |
| AI Refinement | *Start Epic from Jan 1, 2025. Sprint 35-42 | Epic Due Date: May 31, 2025* *Key Result 6:* Refine SignpostAI-powered Zendesk moderation to reduce effort per response (from inquiry opening to completion) by 30%, while maintaining or improving the quality of moderation. This Epic focuses on analyzing and optimizing Zendesk moderation workflows through AI-powered features. The goal is to address bottlenecks, improve efficiency, and maintain the quality of responses, all while gathering feedback to refine tools and processes. *Story Points:* Maximum Points across Developers and Non-Dev Contributors (30–40) across 8 Sprints - 4 months. *Deliverables:* # *Workflow Analysis – End Date: February 28, 2025* * Conduct a baseline analysis of current moderator workflows, including time and effort metrics. * Identify key bottlenecks or repetitive tasks within the moderation process. * Collaborate with moderators to document current challenges and areas for improvement. # *Feature Development – End Date: April 30, 2025* * Develop targeted AI-powered features to assist with repetitive tasks, such as: ** *Auto-tagging:* Automate categorization of client inquiries to save time. ** *Response generation:* Provide context-aware draft responses to moderators. * Collaborate with Zendesk to optimize AI integration for seamless use by moderators. * Test the new AI features to ensure alignment with moderator workflows and client needs. # *Evaluation and Refinement – End Date: May 31, 2025* * Test the new features in pilot environments and collect feedback from moderators. * Measure and validate the reduction in effort per response using predefined success metrics. * Refine features based on feedback and performance data to improve usability and quality. * Publish a final report summarizing: ** Results of the effort reduction. ** Lessons learned from the pilot. ** Recommendations for scaling and future enhancements. *Inclusions Checklist:* * Conduct baseline analysis of current moderator workflows and response times. ✅ * Identify key bottlenecks or repetitive tasks in moderation. ✅ * Collaborate with the Zendesk team to optimize AI integration for moderation. ✅ * Develop targeted AI features to assist with repetitive tasks (e.g., tagging, response generation). ✅ * Test and measure the reduction in effort per response. ✅ * Refine features based on moderator feedback and quality metrics. ✅ *Milestones and Deadlines:* * *Workflow Analysis Complete:* February 28, 2025 * *Feature Development Complete:* April 30, 2025 * *Testing and Refinement Complete:* May 31, 2025 | AI | Medium | 35.29% |
| Map UX | *Start EPIC Sprint 34 - 41.* Gather user and client feedback, to create a roadmap with Google Tags, AND make steady improvements that make service mapping more usable measured by the 200% improvement of service article engagement. |# Investigate and Research Service Map and Services Engagement pre and post redesign/styling * Investigate and Analyze user engagement on the service map before and after Monorepo release * Capture comparison pre and post redesign to determine service map engagement changes * Determine root cause via Analytics to understand and make Tech improvements to the service map to increase user engagement and members promotion to the map * Scope Google map requirements from the Services page # Host Retrospective calls after release of Monorepo to build out an website improvement roadmap * Schedule a call with country teams to gather user and client feedback specifically to the service map and services engagement, and captured in a Retrospective format * Analyze, create and prioritize tickets in Jira in current, future and backlog to develop 3 months roadmap * Design focused enhancements on service mapping, and services engagement in general * Sync and collaborate with S&D to identify and scope additional improvements that were missed in Retrospective, and define the Tech roadmap with S&D's feedback on priority # Implement Google Tags for all defined components for all sites tracked back to PowerBi or Google Analytics as our datasource * Meet and confirm Google Tag components (in case it needs review since we last met) * Develop and Test Google Tags for the agreed components identified on Figma (filters, every category, Readspeaker button, contact channels, and service map), and complete set up for each site * Create user documentation on Google Tags feature, where and how data is collected, and upload onto the Help Center * Share Google Tag functionality with S&D Team as part of team synchronization and alignment. S&D to be made aware of the components teams are capturing * Connect Google Tag components with Analytics and release to country teams on new data * Confirm data and measurement pulled from Google Tag is accurate and correct in our Analytics # Launch improvements and changes to all websites, or specific sites (depending on if it's program specific) * Lead UAT sessions with requester (when 'Ready for QA') to ensure expected performance is accurate and correct before release / made improvement adjustments based on ongoing feedback * Schedule service map and services engagements releases to sites on Tuesday before Sprint ends * Analyze and monitor performance post-release to confirm improved engagement via feedback and Google Tags| | Medium | 0.00% | |
| Monorepo Hypercare | *Start EPIC from Dec 16, 2024. Sprint 34-37.* Launch the Monorepo and create the first satisfactory, Self Sufficient web system Signpost has ever had measured by an improvement in the lighthouse scores in mobile and desktop - 100% performance, 20% Accessibility, 25% SEO performance. # Deploy Monorepo & Directus Configuration Process to all Signpost Websites - *December 19, 2024* ## Communicate scheduled release date and time via email and Teams to all before official release to ensure minimal impact and shared expectations ## Complete Monorepo release to all Signpost sites and all accountable members have access and permissions to make config changes ## Share Signpost Global Announcement after teams sites have gone live with Monorepo to enter Hypercare Phase for 1 month (dedicated time for Dev Team to be all hands on deck to troubleshoot, feature requests are out of scope for backlog) # Improve Lighthouse 'Performance', 'Accessibility', 'Best Practices', and 'SEO' Scores from Adchitects, and Internal Investigations - *January 2025*. ## Identify and report Lighthouse improvements and added to future sprints to be actioned by Developers supported by collected data ## Analyze and establish Lighthouse priority (1) SEO 25%, (2) Performance 100%, (3) Accessibility 20%, and (4) Best Practices ## Dev Team action improvement tickets and confirmed changes via Lighthouse inspection # [Post-Deployment] Host Country Hypercare (Troubleshooting) and Identify and Prioritize Feature Improvements - *January 31, 2025* ## Schedule Weekly / Bi-Weekly Checkin calls with timezone Teams to discuss any issues with their website (Post-UAT) ## UAT Tech members to create Jira tickets in current and future sprints as they see fit according to report ## Close out on Hypercase Phase January 31, and communicate out to the Signpost Team | Medium | 37.00% | |
| AI Ingestion | *Start Epic from Jan 1, 2025. Sprint 35-41 | Epic Due Date: May 31, 2025* *Key Result 5:* Create a comprehensive strategy and roadmap for AI-assisted articles and services content creation across text, audio, and visual formats, culminating in the development of functional prototypes that enhance the pilot knowledge base completeness by 25%. _Measured by the change in the percentage of null fields in bot log “search results” over time._ This Epic focuses on building a strategy and tools for AI-driven content creation to improve the completeness and relevance of the pilot knowledge base. It involves analyzing content gaps, designing prototypes, and implementing solutions across various formats, ensuring measurable improvements in knowledge base performance. *Story Points:* Maximum Points across Developers and Non-Dev Contributors (30–40) across 7 Sprints - 3.5 months. *Deliverables:* # *Content Gap Analysis – End Date: February 28, 2025* * Conduct qualitative interviews with stakeholders to determine the types of queries and formats required by users. * Evaluate existing content gaps using bot log “search results” metrics, focusing on null fields. * Identify areas where search engine functionality can be improved to better align with user needs and reduce null results. # *Tool Development – End Date: April 30, 2025* * Design and prototype AI-assisted tools for creating articles, audio snippets, and visual content. * Enhance the “Request an Article” functionality to allow users to suggest and prioritize content needs directly. * Develop features for oral context note input, enabling users to generate structured content. * Implement tagging functionality to improve the discoverability of content based on inbound user requests. * Test prototypes in pilot environments and collect feedback to ensure usability and alignment with user needs. # *Implementation and Iteration – End Date: May 31, 2025* * Deploy AI-assisted content creation tools in pilot locations. * Collect user feedback and refine the tools to improve accuracy and usability. * Measure improvements in the knowledge base completeness by analyzing changes in the percentage of null fields in bot log “search results.” * Iterate on prototypes to align with evolving user requirements and pilot outcomes. * *Milestones and Deadlines:* ** *Content Gap Analysis Complete:* February 28, 2025 ** *Tool Development Complete:* April 30, 2025 ** *Implementation and Iteration Complete:* May 31, 2025 | AI | Very High | 6.67% |
| AI Onboarding | *Start Epic from Jan 1, 2025. Sprint 35-40 | Epic Due Date: Sept 30, 2025* *Key Result 2*: Develop comprehensive onboarding documentation, pilot plans, and a seamless SaaS-product experience for SPAI to increase user adoption across 10 new use cases. This includes securing API endpoints, creating public-facing tools, and designing training programs to ensure accessibility and readiness for scaling. This Epic focuses on creating a scalable and accessible onboarding experience for SPAI users. It includes the development of technical infrastructure, training programs, and API standards to facilitate adoption and scalability while ensuring usability across diverse humanitarian contexts. *Story Points:* Maximum Points across Developers and Non-Dev Contributors (30–40) across 10 Sprints - 5 months. *Deliverables* # *Technical Infrastructure Development* – End Date: *February 28, 2025* * Protect content injection endpoints and create an easy-to-use method for ingesting external content. * Make the AI agent builder product available in open access with a SaaS account creation onboarding system. * Update agent builder versioning and permissioning to support collaborative workflows. * Create public SaaS system documentation. # *Upskilling and Training Programs* - End Date: *May 31, 2025* * Design and launch an AI fellowship and training program with a structured curriculum. * Conduct monthly bootcamps to train stakeholders on AI product usage. * Create and publish an AI piloting toolkit for Mods/field teams. # *API and SaaS Deployment Standards* - End Date: *July 31, 2025* * Implement API security measures, including authentication, scoping, DDoS protections, and rate limiting. * Finalize backend integrations to enable scalable SaaS adoption across multiple pilot locations. * Test and refine the SaaS onboarding system for deployment readiness. # *Launch and Deployment Support* - End Date: *September 30, 2025* * Roll out the SaaS platform to identified pilot locations. * Monitor onboarding processes and collect user feedback to identify areas for improvement. * Iterate on and refine the onboarding experience based on feedback and analytics. | AI | Very High | 0.00% |
| Summary | Description | Component | Priority | Completion |
|---|---|---|---|---|
| [With S&D Pillar] Implement New Signpost - France (Soliguide) | This epic is to create a new Signpost ZD setup and evaluate/ implement a new website | Implementations | High | 100% |
| Create and deploy new methods for importing new data formats into the Service Map | Using Kobo forms to create sendable links and offline links for forms that will connect to Directus and use AI to read data in unstructured formats to create new service mapping drafts. | Operations and Support | High | 100% |
| AI Chatbot - Phase 6 | * *Pilot Evaluation & Scaling Decisions* (October - Ongoing) * *Milestone*: Data-Informed Decision on P2 and Scaling | AI | Medium | 100% |
| AI Chatbot - Phase 5 | * *Pilot Execution* (August - Ongoing) * *Milestone*: P1 Human-in-the-Loop Pilot Completed | AI | Medium | 100% |
| AI Chatbot - Phase 4 | * *Demo Preparation & Pilot Launch Setup* (July) * *Milestone*: Demo Day & Limited Pilot Launch | AI | Medium | 100% |
| AI Chatbot - Phase 3 | * *MVP Development & Pilot Prep* (May-June) ** {{2024-05-01}}- {{2024-06-01}} * *Milestone*: P1 MVP Bot & Dashboard Ready | AI | Medium | 100% |
| AI Chatbot - Phase 2 | * *Phase 2:* Quality Assessment & P1 Planning (April-May) * *Milestone:* P0 Evaluation Results, Roadmap for P1 | AI | Medium | 100% |
| AI Chatbot - Phase 1 | * *Initiation & Prototyping* (Feb-April) * P0 Prototype Ready for Evaluation | AI | Medium | 100% |
| [With S&D Pillar] Implement New Signpost - Tanzania | Launching new signpost for Tanzania | Implementations | Medium | 100% |
| Product Enhancements - Self-service Empowerment Systems Tools | *Create self-service systems tools that empower support teams to resolve issues and achieve goals autonomously.* 1. Decrease the time to resolution for escalated tickets X% 2. Increase the % of support tickets resolved by support and delivery *Mission Statement* To pioneer scalable technologies for humanitarian good, transforming the way information and digital services reached those in crisis. [https://rescue.app.box.com/integrations/officeonline/openOfficeOnline?fileId=1490255200515&sharedAccessCode=|https://rescue.app.box.com/integrations/officeonline/openOfficeOnline?fileId=1490255200515&sharedAccessCode=] | Innovation | Medium | 100% |
| [With S&D Pillar] Implement New Signpost - Nigeria | Launching new signpost for Nigeria | Implementations | Medium | 100% |
| [With S&D Pillar] Implement New Signpost - Info Digna NYC | Launching new signpost for Info Digna NYC | Implementations | Medium | 100% |
| [With S&D Pillar] Implement New Signpost - Sudan | Launching new signpost for Sudan | Implementations | Medium | 100% |
| [With S&D Pillar] Implement New Signpost - Yemen | Launching new signpost for Yemen | Implementations | Medium | 100% |
| [With S&D Pillar] Implement New Signpost - Libya | Launching new signpost for Libya | Implementations | Medium | 100% |
| [With S&D Pillar] Implement New Signpost - Burkina Faso | Launching new signpost for Burkina Faso | Implementations | Medium | 100% |
| [With S&D Pillar] Implement New Signpost - Mali | Launching new signpost for Mali | Implementations | Medium | 100% |
| [With S&D Pillar] Implement New Signpost - Dominican Republic | Launching new signpost for Dominican Republic | Implementations | Medium | 100% |
| Product Optimization- Transition to "Monorepository" Structure for Signpost Websites | Instead of one repo for each site we will have one repo to rule them all. No need to rebuild and redeploy NPM package for SP base, we can just import from a file in the repo. No need to keep asking devops for new repos. No need to manage permissions separately for each of the subfolders in the repo. Everything in one place *Brand TBD* [~accountid:632e466aa84c7f79c385f6be] | Innovation | Medium | 100% |
| Product Enhancements - Signpost Mapping Hub | [*PRD*|https://docs.google.com/document/d/1TsLVxEaJGS7HiOE2seMNvhD1HxVthrIZXojMZ8Chjkk/edit] Who: As the Product Lead What: We are hoping to build a database with a front end interface either within or outside our existing infrastructure that allows us to take open data from various sources (climate data, open humanitarian data, and our service mapping and data systems) in order to support humanitarian responders and coordinating bodies to better use resources, understand service networks, and see how services layover demographic and environmental trends. As our team continues to build client facing mapping systems, we see a growing possibility to create a multi-layered mapping hub for humanitarian responders and decision makers that combines our closed source data with open source data from a multitude of sources and allows for them to analyze needs gaps and receive automated trend reporting based on a whole of data. Decision Makers are interested in the ability to target Signpost information or services based on relevant environmental indicators from open source data sets, ie. socioeconomic status maps, famine early warning systems, drought/flood data, population movement data. Our team also desires an ability to connect real time regional, local, national news themes to trends to geospatial information. All of this data would allow our team to target who we’re reaching and support decision makers in allocating their resources. In order to do this there are a multitude of questions we must answer: What are our data sources, where do they live, what formats are they in? What infrastructure do we need to be able to import data from various report styles? What is our existing data structure, and what systems infrastructure do we need in order to incorporate open data sources? What will our data stack look like? Should we have one database or a multitude of smaller databases connected via api calls? What automated reporting of environmental data can be done? How do we connect the data in the map to existing processes that impact and support clients? How might we layer data and create filters to give our users the ability to surface the data they need? How will we provide user access visibility to teams with varying needs and data with varying sensitivity? Why: -Create a foundation to build out early action or anticipatory action of crises -Give local partners and partners access to decisionmaking data and coordination information -Expand Signpost’s ability to use geographic information as a means to direct responses or provide information -Expand advocacy possibilities using geospatial information -Support crisis analysis team to standardize more ways to analyze information -Position Signpost to be a leader in emerging climate work | Innovation | Medium | 100% |
| Product Enhancements - LLMAI Chatbot (MVP) | [*PRD*|https://docs.google.com/document/d/1Dv5QH7w2cni2CwIYQ-j8le2PpXcJY-EKgKNInfBUrP8/edit?usp=sharing] Who: As a Project Director What: Select a foundation model LLM and create a human in the loop chatbot in Zendesk with Zendesk knowledge base to offer moderators suggested responses. -To work on any Zendesk connected channel -Will train only on information within Zendesk to begin -Chatbot to use information in our knowledge base and then if that is not clear we train down to moderator responses. -Would be very useful if we have 100 questions, 100 answers for each country -Needs to recognize the country/ localize. Are we making 35 bots or one localizable bot? -Recognize gaps in our knowledge base Must have quality specs for a foundational model -Only provides Accurate/ verified information -Can recognize and respond to the language -Not Unnatural/ uncomfortable language speech or speech patterns -No Creepy/ scary/ or abusive responses -Aware of limitations -Knows not to cause harm Why: [Problem 1] With over 100 million people displaced worldwide, a stark increase in armed conflict, growing incidence of climate-related disasters, and economic hardship, more and more people are forced to make life-changing decisions under duress with limited to no information about their options. Signpost, the largest information project in the aid sector, only serves an estimated 3% of the global population of displaced people. Aid organizations are overwhelmed by the flood of information requests present in every crisis, and the need is growing to up to 1.2 billion people displaced by climate by 2050. Therefore, we need solutions that stretch our limited resources further. [Problem 2] Mis and disinformation spread like wildfire through social media in times of crisis and turmoil. This trend is expected to increase with the mainstreaming of generative AI from both maligned actors and well-intentioned but under-resourced humanitarian agencies. The frequency of updates to information and the size of the Signpost content library pose logistical challenges as moderators are familiar with a limited portion of the information, leading to underutilization of extensive resources and the possibility of citing out-of-date information. [Problem 3] There is already an emergence of unsafe, potentially harmful uses of generative AI as chatbots in the humanitarian sector. Untrained and unsupervised chatbots are being released by private companies with profit incentives and being marketed to dire humanitarian scenarios, putting the most vulnerable people at risk. We want to work with Google.org in this Accelerator to mainstream an alternative that is powered by evidence. [Problem 4] Humanitarian responders deprioritize the creation of humanitarian information where it does not exist and struggle to be responsive to people in need of information. Signpost stands out as the only global-scale, high-quality information service in the aid sector. Currently, when incoming questions reveal gaps in our knowledge, editorial teams manually create content while moderators wait to respond, leading to longer wait times and bottlenecks for clients. The moderation teams are also constrained by time, especially for emergency queries, so they can't always respond immediately, as messages often need extra research for accurate answers and can arrive in waves or unpredictable times, often outside regular business hours. [Problem 5] Signpost has a wealth of untapped natural language data that currently is analyzed through data entry and cleaning from manually submitted forms that only capitalize on a portion of possible insights, making it slower and less effective to disseminate data-driven content. Tagging conversations is also a time-consuming part of the moderator's workload, distracting from more mission-critical objectives. | AI | Medium | 100% |
| Product Enhancements - Cash Distribution (MVP) | Who: As Product Lead What: I need the team to do: - Build out an MVP that includes: 1. digital registration in a new zendesk brand, via Zendesk 2. intake process with customized form that includes the preferences of delivery modality 3. investigate the value and if possible at connections with a cash API in one org or closed consortium (IRC and Mercy Corps only). 4. Create a name, brand, and decks to make a case to get buyin within Mercy Corps leadership 5. Align on possible pilot locations This does not include: 1. development of whole system. 2. buildout and implementation of full pilot 3. creation of a central database of cases to work in consortium (multiple cash disbursing organizations) Why: So Mercy Corps can have a visual and clear concept to begin the process of engaging their programs to implement a digital cash system through Signpost. This will come in three phases (MVP), Pilot, Product. This innovative system would enable secure and efficient cross-border transactions, eliminating traditional barriers and making aid disbursement quicker and more accessible to recipients in different regions. | Innovation | Medium | 100% |
| Product Optimizations - Add Google Tag to Signpost_Base | Add Google Tag to Signpost_Base | Innovation | Medium | 100% |
| Product Optimizations - Research, Analysis, and/or Testing | Product Optimizations - Research, Analysis, and/or Testing * Specific to changing an existing process in order to increase the performance and favourable outcome | Innovation | Medium | 100% |
| Product Enhancements - Research, Analysis, and/or Testing | Product Enhancements - Research, Analysis, and/or Testing * Specific to improving and how we can break through to reach a much higher level of performance for a process that does not yet exist | Innovation | Medium | 100% |
| Implement New Signpost - RAI Virtual Services | Support the RAI Virtual Services team with their deployment of a virtual casework system. | Implementations | Medium | 100% |
| Implement Whatsapper - for Interested Signpost sites | Implement Whatsapper - for Interested Signpost sites | Implementations | Medium | 100% |
| [With S&D Pillar] Implement New Signpost - Thailand | Implement Thailand - new Signpost Site | Implementations | Medium | 100% |
| Implement New Signpost - Uganda | Implement Uganda - new Signpost site | Implementations | Medium | 100% |
| Implement New Signpost - Burkina | Implement Burkina - new Signpost Site | Implementations | Medium | 100% |
| Implement New Signpost - Peru | Implement Peru - new Signpost site | Implementations | Medium | 100% |
| Convert Julisha to the global template | Convert Julisha to the global template | Implementations | Medium | 100% |
| Product Optimizations - Website UX + Style Enhancement FY24 | [*PRD*|https://docs.google.com/document/d/10qNNnDjbtBXtvCpMo0bxmn28LtjPDhokwZiECcNEAyo/edit] *- Deadline for June/July 2024* This is focused on enhancing our Signpost Websites. Determine what improvements we would like to make, have identified, and what we would like to have when comparing our websites with other humanitarian websites. Project folder: [https://rescue.app.box.com/folder/230529781232|https://rescue.app.box.com/folder/230529781232|smart-link] | Innovation | Medium | 100% |
| Convert Sheega to the global template | Convert Sheega to the global template | Implementations | Medium | 100% |
| Convert Importa Mi to the global template | Convert Importa Mi to the global template | Implementations | Medium | 100% |
| Convert Info Palante to the global template | Convert Info Palante to the global template | Implementations | Medium | 100% |
| Convert Info Palante Ecuador to the global template | Convert Info Palante Ecuador to the global template | Implementations | Medium | 100% |
| Convert Italy .Refugee.info to the global template | Convert Italy .Refugee.info to the global template | Implementations | Medium | 100% |
| Convert Simaet Bhatha .com to the global template | Convert Simaet Bhatha .com to the global template | Implementations | Medium | 100% |
| Convert Bolo-pk.info to the global template | Convert [http://Bolo-pk.info|http://Bolo-pk.info|smart-link] to the global template | Implementations | Medium | 100% |
| Implement New Signpost - Bangladesh | Implement Bangladesh - new Signpost site | Implementations | Medium | 100% |
| Convert Signpost-Test to the global template | Convert Signpost-Test to the global template | Implementations | Medium | 100% |
| Convert Settlein .support to the global template | Convert Settlein .support to the global template | Implementations | Medium | 100% |