Back to portfolio
Data Extraction · Cross-functional

Stakeholder Tension Mapping Engine

Stakeholder Tension Mapping Engine

Messy stakeholder interview data → extracts competing priorities and maps them on continuums → reveals structured tension landscapes for decision-making.

01The Problem

Organizations often collect stakeholder input but struggle to reconcile conflicting perspectives, definitions, and priorities. Without a structured way to surface these tensions, important differences get flattened or ignored, leading to misalignment and poor decisions. The difficulty is not lack of data, but lack of clarity about how viewpoints diverge.

02What the AI Does

I analyze stakeholder interview transcripts to extract perspectives, classify priorities, and identify competing viewpoints. I structure these into continuums, map stakeholders across those spectrums, and generate visual and narrative tension-spectrum models. I compare language, definitions, and assumptions across participants to surface both explicit disagreements and subtle conceptual differences. I am a configured instance of ChatGPT (GPT-5.3) with no external tools or retrieval systems, operating purely through prompt-based instruction and structured output formatting.

03Design Decisions

01 · Choice

Narrow specialization on stakeholder interview analysis and tension mapping

Why

To prioritize depth and rigor in a specific analytical task rather than broad general-purpose assistance [Creator: add rationale]

Constraint

Prevents drift into generic summarization or unrelated advisory tasks

02 · Choice

Use of a proprietary spectrum-based tension-mapping framework as the core analytical structure

Why

To shift from binary disagreement framing to spectrum-based understanding of competing priorities [Creator: add rationale]

Constraint

Forces outputs to represent nuance and gradients instead of oversimplified positions

03 · Choice

Explicit requirement to capture micro-level differences (language, terminology, definitions)

Why

To reveal hidden sources of misalignment that are often overlooked in high-level summaries [Creator: add rationale]

Constraint

Increases analytical depth and prevents superficial synthesis

04 · Choice

Separation of mapping tensions from solving them

Why

To maintain objectivity and avoid premature convergence on solutions [Creator: add rationale]

Constraint

Disallows recommendations or facilitation outcomes within outputs

05 · Choice

Structured output format including continuums, mappings, and synthesized insights

Why

To make complex qualitative data legible and comparable across stakeholders [Creator: add rationale]

Constraint

Outputs must be organized and systematic rather than freeform narrative

06 · Choice

Instruction to document both explicit and implicit assumptions

Why

To surface underlying mental models driving stakeholder positions [Creator: add rationale]

Constraint

Requires inference but within grounded evidence from transcripts

07 · Choice

No external tools, data sources, or memory integration

Why

Simplicity and control over inputs/outputs; ensures analysis is fully contained within provided transcripts [Creator: add rationale]

Constraint

Cannot validate claims or enrich context beyond given material

08 · Choice

Emphasis on comprehensive coverage over brevity

Why

To ensure no relevant tension or perspective is omitted [Creator: add rationale]

Constraint

Outputs may be dense and require interpretation by a human facilitator

05Key Insight

AI is most effective in stakeholder work when it structures disagreement, not when it tries to eliminate it.