ProgramSuite

Streamlining campaign management with a unified platform for efficiency, transparency, and collaboration

Organization: Color Of Change

Time to launch:

12 months

Collaborators:

40+ (UX designers, researchers, software engineers, data analysts, campaign managers, communications, project managers)

  • Campaign teams were using 12+ disparate tools, causing inefficiencies, lack of transparency, and disconnected workflows. This fragmentation hindered cross-collaboration and created redundant processes that slowed down the organization's ability to effectively drive campaigns.

  • A unified ecosystem that consolidated campaign management workflows, enhanced transparency, reduced repetitive tasks, and facilitated better planning and collaboration across teams. The platform was designed specifically to empower campaigners by leveraging technology and data for member-centered solutions.

    • Increased ease of use for campaign management

    • Higher adoption rates across the organization

    • Reduced time spent on repetitive tasks

    • Improved cross-team collaboration

    • Enhanced transparency in campaign planning and execution

    • Centralized campaign planning and management

    • Cross-functional collaboration tools

    • Streamlined workflows

    • Integrated data reporting

    • User-friendly interface based on extensive user research

    • Transparent project tracking

    • Led design sprints and synthesized 500+ inputs into core themes and features

    • Managed cross-functional team of 70+ collaborators

    • Established representative user groups of 20+ for continuous feedback

    • Orchestrated 10 monthly showcases and 3 interactive trainings

    • Idea validation & need identification

    • Product strategy & roadmapping

    • User journey mapping

    • Product requirements engineering

    • Python

    • Django

    • JavaScript

    • NextJS

    • AWS

    • Terraform

    • PostgreSQL

    • BigQuery

    • GitHub

    • JIRA and Asana (project management)

Crisis Contact Simulator

Empowering counselors with AI: ensuring every word counts in crisis scenarios

Organization: The Trevor Project

Time-to-launch:

6 months (March 2021)

Collaborators:

40+ (machine learning scientist and engineers, UX researchers, designers, software engineers, data scientists, product managers, Internal Education & Clinical Services)

  • With the ambitious aim to serve 1.8 million youth, The Trevor Project recognized a need to expand its counselor base to manage crisis interventions effectively. However, the counselor training process faced limitations due to its manual and resource-intensive nature, hampering the scaling of the counselor workforce. The challenge lay in enhancing the training capacity without compromising the quality and efficacy of the training, ensuring counselors are prepared to navigate real-life crisis scenarios involving LGBTQ youth.

  • A machine learning-powered platform that automates and amplifies the counselor training process. CCS immerses trainees in realistic and relevant conversations with a simulator, programmed to emulate an LGBTQ youth in a crisis scenario. This not only ensures that the counselors are proficiently trained but also allows for the scaling of the training cohort, thereby increasing the number of graduating counselors.

    • NLP conversational model

    • Simulated crisis conversations

    • Realism & relevance in training

    • Integration into chat & text platform

    • In-App survey for continuous improvement

    • Managed product lifecycle from conception to launch in 6 months, on time and within budget

    • Designed simulator that achieved 94% relevance and 89% topical accuracy ratings from trainees

    • Implemented iterative learning approach showing 74% of trainees reported solid understanding after second roleplay, compared to just 43% after first roleplay

    • Created training solution where 72% of trainees cited simulator practice as the most valuable component

    • Conducted post-launch analysis showing 211% increase in counselor training capacity and 23% improvement in counselor preparedness scores

    • Featured in TIME's Best Inventions of 2021, generating significant positive PR and securing $1.2 million in additional funding opportunities

    • Orchestrated UAT with 15 training coordinators and led pilot program with 20 trainees

    • Google Cloud Platform

    • GPT3 (fine-tuned on internal data)

    • BigQuery

    • Salesforce

TrevorSpace AutoMOD

Enabling content moderation and timely crisis intervention online

Organization: The Trevor Project

Time-to-launch:

6 months (July 2021)

Collaborators:

15+ (machine learning scientist and engineers, UX researchers, designers, software engineers, data scientists, product managers, operations, Clinical Services & Research)

  • The urgent challenge of swiftly identifying and addressing potentially suicidal content among posts from LGBTQ youth on TrevorSpace requires immediate and effective response mechanisms. The human moderation team, while diligent, faces the enormous task of reviewing massive volumes of posts promptly, thus raising the potential risk of delayed intervention in critical moments.

  • Utilize specialized algorithms, trained with actual content from TrevorSpace, to automatically analyze all user posts and flag possible suicidal content for instant review by a human moderator. This AI-assisted approach not only accelerates the identification and response to critical content but also ensures that human moderators can intervene more quickly and effectively, providing crucial support when it's most needed and reinforcing a safety net for the youth interacting on the platform.

    • Response time

    • Accuracy of flagging

    • Escalation rate

    • Real-time post analysis

    • AI-powered content flagging

    • Human moderator alert

    • Crisis intervention workflow

    • Spearheaded ideation sessions that identified critical need for automated moderation, resulting in implementation that achieved 30% improvement in harmful content detection

    • Implemented system that analyzes 100% of text-based posts, significantly reducing emotional burden on youth

    • Directed design implementation with 3 UX iterations, reducing moderator workflow steps by 20%

    • Led user testing with 3 moderators, incorporating feedback into final product

    • Google Cloud Platform

    • BigQuery

    • TrevorSpace

Risk Assessment Analysis

Intelligently prioritizing conversations for timely crisis support

Organization: The Trevor Project

Time-to-launch:

6 months (July 2020)

Collaborators:

15+ (machine learning scientist and engineers, UX researchers, designers, software engineers, data scientists, product managers, Clinical Services & Research)

  • The critical issue addressed revolves around efficiently and effectively prioritizing and managing the communication with LGBTQ youth reaching out to The Trevor Project’s digital platform for crisis assistance. The complexity arises from moments when the volume of youth requiring help surpasses the available counselors, creating a need to immediately identify and prioritize those at imminent risk of harm to reduce wait times and expedite connection to vital support.

  • Leverages the potency of Natural Language Processing and machine learning to automate and refine the crisis contact intake process. Operating on a binary text classification system, it analyzes initial communication and intake information from the youth—utilizing answers to crucial questions and applying intelligent assessment criteria to categorize them into standard and priority queues.

    • Wait time for high-risk youth

    • False positive & negative rates across 20 demographic/intersectional categories

    • Intelligent prioritization

    • Binary text classification

    • Inclusiveness & fairness metrics

    • Comprehensive risk assessment

    • Testing & feedback integration

    • Deployment & implementation

    • Impact measurement & reporting

    • Dashboard visualizations

    • Google Cloud Platform

    • ALBERT (fine-tuned on internal data)

    • BigQuery

    • Looker Studio

    • Salesforce

myLanguage

Breaking language barriers with real-time translations

Organization: myLanguage

Project length:

12 months (enhancements to live product)

Collaborators:

15+ (ML engineers, UX designers, language experts, hardware and software engineers)

  • In a globalized world, clear and immediate communication across language barriers is essential. Traditional language translation apps have struggled with providing accurate, contextually correct translations in real-time during live interactions. Most available solutions are reliant on internet connectivity, the lack of offline functionality in most translation apps restricts their usability in various practical scenarios and locations, limiting users in their ability to communicate freely anywhere, anytime.

  • Live-translated conversations without the awkward pauses or misinterpretations. The platform translates voice and text communications accurately and efficiently and can be trained to comprehend any language, dialect, and domain. Prioritizing user privacy and convenience, all translations are processed locally on the user's device, eliminating the need for data to be transmitted through the cloud and enabling secure and efficient conversations.

    • Translation accuracy

    • Product usage

    • Client feedback

    • AI-powered translation

    • Real-time translation

    • Offline/online capability

    • Conversation mode

    • Language detection

    • Product strategy

    • Roadmap planning

    • Vendor governance

    • System analysis

    • Testing

Additional Work

STP Connectors

Automated relationship-driven sales platform leveraging warm introductions to streamline lead tracking, follow-ups, and revenue growth.

Sherpa

Automated project execution platform optimizing real estate workflows, task management, and reporting

Rover

Real estate pipeline management system streamlining deal sourcing, due diligence, and opportunity tracking

Agile Team Enablement

Provided Agile training and team enablement support, equipping teams with frameworks and best practices to improve collaboration, accelerate delivery, and enhance adaptability in AI-driven defense and security solutions.

Connect

Streamlines, standardizes and automates real-time communication and workflow between clients and PwC Engagement teams
(~600k users; 1M expected by 2020 year end)

Payment Gateway

Retail & eCommerce payment platform processing $1B annually

Ashley eCommerce store

Salesforce Commerce Cloud migration; including APIs & file-based integrations

Determine
(acquired by corcentric)

Complete Source-to-Pay solutions with supplier management, sourcing, CLM, procure to pay and analytics solutions

Virtual Check Deposit

Integrating OCR technology to mobile apps for check deposits

ApplePay & GooglePay

Client enrollment & implementation support for mobile wallet solutions

Card Fraud Alerts

Automated detection & notification for fraudulent activity