Emmanuel Boateng
Home · § Projects

Tools that
take their side.

Research earns its keep when it ships. These are the working systems: conversational agents and machine-learning models, built for the people inside complex projects: their safety, their wellbeing, their work.

§ 01 · Applied work ↓4 builds · 2 awards · since 2022
§ 01 · Applied work

Four builds

since 2022
PsycheGuard interface
01Flagship · award-winning

PsycheGuard

A confidant in your pocket, for the people on site.

An AI-powered conversational app that helps construction workers understand and improve their psychosocial wellbeing: a private, judgement-free first step toward support in an industry where reaching out is hard. Built as an agentic system with a safety-first conversation design.

Conversational AIVoiceflowPsychosocial safetyConstruction
Social Impact Award + People’s Choice Award · UNSW Digital Makers Competition, 2024UNSW Founders Micro Grant · with UNSW & Queensland University of Technology
Recognition
Receiving the award · UNSW Founders, 2024
Receiving the award · UNSW Founders, 2024
Feedforward Agent interface
02Prototype · showcased

Feedforward Agent

Feedback that points forward, at scale.

An AI agent that helps academics give students feedforward: guidance aimed at the next attempt, not just a grade on the last one. Designed to make high-quality, personalised assessment guidance feasible across large cohorts.

AI agentAssessmentHigher educationLangChain
Showcase talk · “The Case for AI Agents in Assessment Feedforward” · AI & Assessment Symposium, UNSW Canberra, June 2026Learning & Teaching Innovation Grant (A$14,630) · University of Wollongong
Safety Knowledge Chatbot interface
03Prototype · showcased

Safety Knowledge Chatbot

Site safety know-how, on demand.

A chatbot for sharing construction safety knowledge: turning scattered guidance, regulations and hard-won site experience into answers a worker or supervisor can get in plain language, in the moment.

NLPKnowledge retrievalWHSVoiceflow
Innovator Showcase · 23rd World Congress on Health & Safety at Work, Sydney, 2023Global Research Small Grants Scheme (A$7,850) · Chief Investigator · University of Wollongong
Recognition
Innovator Showcase · 23rd World Congress on H&S at Work, Sydney
Innovator Showcase · 23rd World Congress on H&S at Work, Sydney
Green-Building Predictor interface
04Industry-adopted

Green-Building Predictor

Cost & schedule certainty for sustainable builds.

A machine-learning model that predicts the cost and duration of green-building projects with far greater accuracy than conventional estimating, giving developers the confidence to commit to sustainable design.

Machine learningCost predictionGreen buildingPython
Implemented in industry by Yau Lee Construction Company LimitedHong Kong Polytechnic University Research Grant (~A$116,863) · Co-Investigator
§ ½ · In the margin
The risk of failure is the price of admission on the path to success.
Emmanuel B. Boatengon risk, initiative & the work that compounds
§ 02 · Toolkit

How these get made

research question → deployed tool

One stack carries a project from a question to something people can use: modelling and text mining in R and Python, agents orchestrated and observed, then shipped to the web.

Modelling & data
RPythonTypeScriptMachine LearningNeural NetworksNLPSentiment AnalysisSocial Network AnalysisSPSSNVivo
AI & agent development
LangGraphLangChainVoiceflowLangSmithLangFuseNext.jsGoogle Cloud RunStreamlit
Construction & project delivery
Primavera P6MS ProjectVico OfficeBIM 4D/5DJiraLinear
Automation & teaching
Make.comn8nZapierCanvasMoodleMiroMentimeterH5P
§ · Correspondence

Always glad to hear from students, collaborators, and editors. Start a conversation →

Office

School of Engineering & Technology

UNSW Canberra · Northcott Drive, Campbell ACT

Send a message →
Profiles
UNSW staff page ↗Google Scholar ↗ORCID 0000-0002-6434-0085 ↗
Elsewhere
LinkedIn ↗X ↗

AIPM · CIOB · GIOC

Explore
HomeAboutResearchProjectsContact
© 2026 Emmanuel B. BoatengPrivacy · Terms · RSS