REMTA WORKFORCE RESILIENCE SYSTEM (AmadursInsight)
One of the biggest challenges we faced during the hackathon was balancing multiple responsibilities while working effectively as a team. Throughout the competition, we had to juggle work commitments, family responsibilities, personal obligations, and the demands of the hackathon itself. Managing these competing priorities required significant discipline, time management, and resilience. For us, winning would be a reflection of the power of teamwork. Our team consists of individuals with different backgrounds, perspectives, and strengths. Despite balancing work, family responsibilities, studies, and personal commitments, we remained focused on achieving our objective. The process required open communication, mutual respect, and a willingness to support one another through periods of uncertainty and changing project directions. Winning would demonstrate that when people collabora
Voting is not currently open.
Project Information
Project Name
REMTA WORKFORCE RESILIENCE SYSTEM (AmadursInsight)
One-Line Problem Statement
Job redundancy fear is the wrong frame: the routine 70% moves to you as a review skill, not away from you as a task. Line-by-line verification against source notes is now the entry-level craft.
Project Description
Generative AI now drafts BRDs and Given/When/Then stories in seconds, work that consumed 60–70% of a junior or mid-level BA's time. That time doesn't disappear. It moves to the three things AI still can't do: designing the extraction contract, facilitating the humans, and deciding what's good enough to ship. 0 of that time is reclaimed automatically. Who signs off on what? Role: Junior BA Reviews: Every requirement, story, and criterion line by line against the source notes Before it may proceed: Marks each item reviewed; corrects wording in place (the tool's own edit step) Role: Senior BA Reviews: The Conflicts section specifically, plus a spot-check of Assumptions and Open Questions Before it may proceed: Confirms each conflict has been routed to the named resolver, not resolved by guesswork Role: BA Practice Lead Reviews: The extraction prompt and schema themselves (governance, not individual documents) Before it may proceed: Owns what "reviewed" means org-wide; any prompt change
Methodology
Unvetted AI output introduces specific, nameable risks: hallucinated edge cases, missed non-functional requirements, unaddressed stakeholder conflicts, and no clear owner of accountability. This protocol describes the six mechanisms REMTA Workforce Resilience System (AmadursInsight) uses to keep a human answerable for every requirement before it reaches a backlog and names where those mechanisms currently fall short. Nothing produced by this prototype is treated as approved until a named human has signed off at the level above. The tool has no concept of "approved"; that status lives outside it, in whatever system of record the organization already uses. Named limitations: Mechanism 02 reduces invented detail by instruction; it is not an independent hallucination check. A second, differently prompted model pass to cross-examine the first draft would be a stronger guarantee and is not yet built. Conflict detection (03) only sees what is pasted into the backlog field. It does not connect live to Jira, so a conflicting ticket the reviewer forgets to paste in is invisible to it. The protocol assumes the three sign-off roles above are staffed and distinct. In a team of one, the same person plays all three, which is workable but removes the second-reviewer effect the table is designed to give. Compliance record-keeping (who reviewed what, when) is a process today, not a logged feature of the tool. Formalizing it as an audit trail is the natural next build.
Team Bio
Our team is a diverse group of multicultural and multilingual professionals united by a shared passion for innovation, problem-solving, and continuous learning. We bring together different cultural perspectives, professional experiences, and skills
Last updated just now
…