← All work

Case study — 2020 — 2025

BRIDGR — Client Evaluation Platform

BRIDGR digitizes the client evaluation process for consultants: automated assessments replace paper-based workflows, a scoring system rates each client, and an analytics dashboard with exportable reports supports decision making. I introduced AI in two targeted places — summarizing client answers for consultants, and drafting recommendations they review before sending.

BRIDGR evaluation dashboard showing organization score of 76.8, a bar and line chart comparing team scores against benchmarks across five supply chain dimensions, and entry points for AI insights and delegated response
fig. 01a — The evaluation dashboard consultants use to track client scores against industry benchmarks across every assessment dimension.
BRIDGR AI-generated evaluation insight panel showing a written summary of the client's sustainability assessment broken down by dimension, with options to select insight language and download an editable report
fig. 01b — The AI-generated summary that replaces manual read-through of client answers, giving consultants a concise starting point instead of raw responses.
BRIDGR detailed analysis view showing score and level, 66% team alignment, a team distribution donut chart by sustainability level, and a table breaking down scores and alignment rate per evaluation dimension
fig. 01c — Detailed analytics giving consultants per-dimension scoring and team alignment data to support their recommendations.
BRIDGR global dashboard for the Supply Chain Sustainability Assessment questionnaire showing 10 organizations, 49 users, average score of 70.7, score distribution pie chart, and evaluations created over time
fig. 01d — Aggregate view across all client evaluations for a given questionnaire, used to track program-wide performance, not just a single assessment.
BRIDGR questionnaire screen showing Question 4 of the Strategy and Leadership section, a multiple-choice question about completed digital transformation projects, with a 4% progress bar at the bottom
fig. 01e — The evaluation experience end users complete, feeding the data the AI summarizes and consultants act on.
AI features cut the time consultants spend per client by an estimated 80%, based on before/after observation.

The problem

Consulting firms evaluated clients through a manual, largely paper-based process: collecting answers, reading through every submission, scoring by hand, and writing recommendations from scratch or from memory. Each client took hours of repetitive work before the consultant could get to the judgment calls that actually needed their expertise.

The platform

BRIDGR digitizes and automates the assessment phase end to end. Clients complete structured evaluations online; a scoring system rates them against defined criteria; and a detailed analytics dashboard surfaces the metrics consultants need for decision making, with exportable reports for stakeholders.

A recommendation engine then defines and generates recommendations for each client, closing the loop from assessment to actionable advice.

AI, in two deliberate places

Rather than sprinkling AI across the product, we introduced it where it removed the most manual work. First, evaluation: the AI reads a client's submitted answers and produces a concise summary, replacing the consultant's manual read-through of every response. Second, recommendations: the AI generates a draft recommendation from the summary and the client's stats. The consultant reviews and edits that draft before anything reaches the client — it is a starting point, never an auto-send.

I owned the architecture of the AI feature implementation end to end, integrating LLM capabilities via Vertex AI into the existing platform.

My role

Senior Software Engineer on a remote team (Montreal-based company, 2020–2025). Beyond the AI work, I designed and scaled SaaS features — dashboards, RESTful APIs — owned frontend and backend architecture on a number of tasks, and mentored and led two developers through code reviews, CI/CD implementation, and sprint planning.

Outcome

The AI features cut the time consultants spend completing all phases with one client by an estimated 80% — an observed estimate from comparing time-to-complete before and after the features shipped, not a formally instrumented metric. The summarization and drafting features were adopted into consultants' daily workflows.

Want the full story behind this project?

houssem.djeghri@gmail.com