Mark Z Gleason

Playbook / 01

AI-Assisted
Technical
Discovery

Reduce discovery loops by giving product workflows system-level context before engineering refinement begins.

System-Aware Product Discovery For Complex Platforms

Last updated: 2026-05-17

PRD Framework
↓
Repo Knowledge Layer
↓
Architecture References
↓
AI Discovery Workflow
↓
Scoped Technical Questions
Impacted Systems & Dependencies
↓
Engineering Refinement

The Problem

Discovery slows when
system knowledge becomes
organizational memory.

Implementation knowledge is distributed across repositories, teams, and historical decisions. Product managers often rely on synchronous engineering conversations to uncover edge cases, integration points, and downstream impact - creating bottlenecks and context switching that slow everyone down.

Edge Cases

Integrations

Tribal Knowledge

Cross-Team Dependencies

Implementation Uncertainty

A Discovery System That Builds Context

Move from idea to a scoped, technically-aware plan-faster.

01

PRD Framework

Start with a proven structure. Define objectives, constraints, and success metrics.

02

Repo Mapping

AI summarizes each repository, its purpose, responsibilities, and primary connections.

03

Architecture References

Build a layered understanding of systems, services, and data flows.

04

Discovery

AI asks targeted questions, identifies edge cases, and uncovers assumptions.

05

Implementation Awareness

Surface likely change areas, files, and impacted systems.

06

Engineering Review

Collaborate with engineering to refine, validate, and finalize the plan.

✦

The result: better first-pass PRDs, fewer discovery loops, and higher quality engineering conversations.

Repository Summary

loan-application-service

Purpose

Core service for creating and managing loan applications from intake to decisioning.

Primary Connections

  • lead-ingestion-service
  • customer-identity-api
  • underwriting-engine
  • document-service

Common Change Areas

  • validation/
  • workflow-routing/
  • decision-logic/
  • integration-mappers/
  • api/

System Relationships

System relationships map for loan-application-service and connected services

Likely Change Locations

src/workflows/

src/validation/

src/integration/mappers/

src/api/

Considerations

  • Impacts underwriting flow
  • Triggers document generation
  • Affects customer notifications

Confidence

High

What This Does Not Replace

This does not replace engineering judgment.

  • Engineering judgment
  • Architecture review
  • Production validation
  • Implementation ownership
  • Technical tradeoff analysis

Failure Modes

  • Stale architecture context
  • Hallucinated dependencies
  • Outdated implementation paths
  • Over-confidence in AI output
  • Missing organizational context

Outcomes

  • Better first-pass PRDs
  • Faster discovery cycles
  • Earlier edge-case visibility
  • Reduced tribal knowledge dependency
  • Higher-quality engineering discussions

Build the system. Compound the advantage.

This playbook helps product and engineering teams turn fragmented knowledge into clarity, alignment, and momentum.

Let's Work Together View All Playbooks