Skip to content

Kellogg AI Lab · Applied AI Lab · Northwestern

Building a collaborative ecosystem

The Build Studio

Everything being built here is visible and joinable. Browse live projects across every theme and stage, then take a role, or bring a problem and start one.

Start a project2 open calls
Theme
Stage

13 projects

Launched
Marketing

Intro-call researcher

Reps spent ~15 minutes hand-researching every prospect across CRM, LinkedIn, and the web (~34 hours a month). This agent compiles a full pre-call briefing from internal and web sources in about a minute, 93% less prep time and ~32 staff-hours back every month.

View
Launched
Marketing

Inbound-lead qualifier

Reps hand-qualified every inbound demo request (~33 hours a month). This agent enriches and scores each lead against the ideal-customer profile and posts a decision into Slack, ~27 rep-hours freed every month at 95%+ accuracy.

View
Launched
Operations

Legal document drafter

A lean legal team hand-drafted NDAs, addendums, and standard agreements (~5 hours a month on template work). This agent auto-drafts four document types, enriches missing company data, files each in the right place, and routes only true exceptions to legal, under a minute and ~$0.12 a doc, with review cut from ~5 hrs/month to under 30 minutes and 0% misfiled.

View
Launched
Strategy

Sales business-case writer

Account execs spent hours hand-compiling business cases from call notes and CRM data. This agent drafts a full, editable business case from just a company name, pulling call transcripts, CRM data, and web research, in about 5 minutes instead of hours, saving ~12 hours per rep every month.

View
Launched
Operations

Internal-operations agents

HR, finance, recruiting, and sales ran on manual workflows across siloed SaaS tools. The team built a suite of agents, pipeline intelligence, research briefs, pitch decks, interview guides, cash-flow reporting, wired to the company's own systems. The CEO cut C-suite meeting prep 80–90% using their output.

View
Launched
Marketing

Customer-proof extractor

A lean marketing team hand-combed call recordings for quotes and case studies (~6 a quarter). This agent scans transcripts for proof points, ranks them against current priorities, and drafts outreach and interview guides, an 87.5% hit rate finding case-study-ready calls across 24 customer accounts.

View
Launched
Operations

Natural-language dashboards

Building a custom dashboard took technical skills non-technical staff didn't have. This agentic workflow assembles a complete dashboard from a single plain-English question, a question in, a built dashboard out.

View
Launched
Finance

Trend-vs-anomaly analyzer

A metric drop could be a one-off blip or a multi-year trend, and leaders couldn't tell which. This agent classifies each movement as trend or anomaly across years of data, with confidence flags, so leaders act on the patterns that matter.

View
Launched
Strategy

"State of the org" storyteller

Leaders spent 1–3 hours manually gathering data and formatting slides for stakeholder presentations. This guided wizard turns an organization's data into a board-ready slide deck in under 10 minutes.

View
Launched
People

AI seller-training voice coach

Reps had no scalable, consistent way to practice discovery and competitive-selling calls. This always-on AI voice agent lets reps role-play a live sales call against five dynamic buyer personas (CFO, RevOps, IT leader, and more) for any target company, ending each session with rubric-scored coaching and specific next-time priorities.

View
Launched
Marketing

Customer-deck generator

AEs spent up to an hour building exec-summary, demo, and pricing decks for each deal. This agent auto-generates all three deck types from branded templates populated with CRM and call-intelligence data, drafted in minutes instead of built by hand.

View
Prototype
Marketing

Sales email writer

Reps manually wrote prospecting, re-engagement, and follow-up emails, leaving coverage gaps. This agent drafts lifecycle-aware emails from CRM data, call transcripts, and deal activity as ready-to-edit drafts, so reps keep more of the pipeline warm.

Seeking ML engineer
Pilot
Operations

Bring-your-own-AI data gateway

Enterprise clients wanted to query proprietary data with their own AI tools but couldn't connect securely. This working gateway, with an OAuth identity layer, lets a client's own AI agent query the platform's data in natural language, proving out "bring your own AI" with no raw-database access.

Seeking Backend engineer

Shipped project outcomes shown here are drawn from the Applied AI Lab · Corporate Partnerships Project Catalog, Kellogg School of Management, Northwestern University, a sample of the 28 problems Lab teams have already solved with partners.