pro
Emman Ermitaño
AI Systems Engineer | Automation, RAG, Internal Tools
Ready for work
Emman is ready for their next project!
Stockpile is an AI inventory workflow system built to automate stock operations with tighter control over writes, costs, and long-running tasks.
I built it to validate each action before execution, gate AI behavior through explicit rules, and handle inventory workflows in a way that stays fail-proof in production.
The system was built with fail-closed outputs, auto-repair paths, and async processing so the workflow stays stable under repeated use and heavier job volume.
0
67
Cognitive Security Engine is an AI contract review system built to turn dense legal documents into clause-level risk signals and edit-ready recommendations.
I built it to break contracts into reviewable units, anchor findings to the source language, and score risk against the actual exposure in the document.
The system was built with traceable outputs, safe degradation, and throttling controls so the workflow stays stable under load.
0
53
Highground is an AI market intelligence system built to turn competitor websites into structured intelligence.
I built it to extract the right fields first, route harder inputs through the right processing path, and return output in a format that is easier to review and compare.
The system was built with HTML cleaning, schema-checked output, and quota controls so the workflow stays stable under repeated collection.
0
45
Commercial Proposal Engine is an AI proposal system built to generate finance-grade commercial documents with consistent ROI logic and clean structure.
I built it to separate calculation logic from drafting, enforce the underlying schema and template, and keep proposal generation aligned with the commercial rules behind the output.
The system was built with strict parsing, retry controls, and spend limits so the workflow stays repeatable under client volume
0
42
Sentinel is an AI energy intelligence system built to turn live telemetry into structured answers users can act on.
I built it to convert natural language questions into executable query logic, calculate verified aggregates against live energy data, and return accurate outputs.
The system was built with server-side aggregation, spend controls, and bounded execution so the workflow stays reliable, reviewable, and stable under load.
0
33
Clearworth is an AI financial workflow system built to turn natural language requests into structured finance actions with auditable calculations.
I built it to interpret user intent, route requests into the right operation, and compute totals through deterministic logic before any result is returned.
The system was built with anti-loop safeguards, spend controls, and bounded execution so the workflow stays stable, reviewable, and predictable in scale.
0
34
Arcadia is an AI lesson planning system built to generate structured classroom materials under real pedagogy constraints.
I designed it to handle curriculum inputs, apply rule-based instructional logic, and return outputs in a consistent format that teachers can actually use.
The system was built with validation layers, safety limits, and controlled generation so the workflow stays usable under production conditions.
0
37
Blackletter is an AI legal research system built to generate cited legal memos from live US primary source retrieval.
I designed it to pull authority directly from US legal sources, structure the research flow, and return analysis with traceable support attached to each claim.
The system was built with source-linked outputs, retrieval controls, and separated authority and analysis layers so lawyers and paralegals can review support fast and draft with tighter control.
0
37
MedIntel is an AI research system built to turn live PubMed retrieval into structured clinical briefs.
I designed it to search current medical literature, extract relevant findings, and return outputs in a consistent format that supports faster review.
The system was built with structured extraction, rate limits, spend controls, and deterministic fallback logic so the workflow stays stable under load.
0
42
Brand Strategy Engine is an AI brand positioning system built to turn raw business inputs into structured strategy outputs.
I designed it to process brand context, competitor references, and positioning tradeoffs, then return clear writing rules and strategic direction in a consistent format.
The system was built with schema checks, sanitization, and controlled output logic so the results stay usable across real working drafts and downstream content workflows.
0
43