A Genuinely Different Way of Running Events

The Challenge

Algebra Consulting founder Waleed Jahangir was already using ChatGPT, but on an ad-hoc, single-chat basis — every proposal written from scratch, every conversation starting from zero context on the organisation, its past events or its partners. Nothing persisted between sessions, so nothing compounded: nothing built on the work that came before it. Event proposals and planning decisions still relied on manually reassembling the same background information each time, which meant AI was speeding up typing without changing how the work actually got done.

The Approach

A full 1:1 walkthrough of Claude, not a single-feature demo — moving from ad-hoc, single-tool habits to a structured setup covering:

• Memory — giving Claude persistent context on the organisation, so it stops being re-briefed from scratch every session
• Skills — packaging repeatable tasks (like proposal writing) into reusable, reliable workflows
• Projects — organising ongoing work — event cycles, partner relationships — in one place instead of scattered chats
• Routines and schedules — automating recurring work so it runs on a cadence instead of being manually triggered every time
• Cowork, for ongoing hands-on build work rather than one-off prompting

Built around Waleed's actual proposal and event-planning documents throughout, not generic examples.

The Systems Built

1. Automated Proposal Assistant

Built as a Claude Skill trained on Algebra Consulting's own past proposals, not a generic template. Fed a short brief — the partner, the event, the ask — it produces a first-draft sponsorship or partnership proposal already in the organisation's tone and structure, with the commercial terms it tends to offer built in. Waleed edits and sends, rather than starting each one from a blank page.

2. Connected Tools & Live Dashboards

Rather than each system living in its own silo, the underlying tools are connected, so information flows between them instead of being copied by hand. For one event, inbound stall enquiries came in through a form and normally sat in a Google Sheet, checked manually. Using Claude's Artifacts, that became a real-time dashboard instead — enquiries visible at a glance as they landed, with the option to ask the AI direct questions about that data rather than scanning rows by hand.

3. On-the-Day Operations & Partner Documentation

Built on Projects, using the meeting transcripts Algebra Consulting already had rather than starting from nothing. Raw planning-meeting transcripts get turned directly into a structured on-the-day operational plan — running order, responsibilities, contingencies — and into partner-facing documentation, generated from the same source material instead of written up separately by hand. As much of this was about teaching Waleed how to think and work with AI as it was about the systems themselves — a shift in mindset that outlasts any single document.

The Outcomes

• Waleed is no longer the bottleneck for proposals and event admin — first drafts go out in a fraction of the time they used to take him personally to write from scratch
• Algebra Consulting now runs on one connected picture instead of scattered spreadsheets, chats and documents rebuilt for every event
• Stall enquiries are visible to Waleed live, as they land, instead of sitting in a spreadsheet he had to check manually
• His own role shifted from doing the admin to directing systems that do it for him — freeing up time and headspace for the judgement calls only he can make
• On-the-day plans and partner documentation now come from the same source material every time, so quality no longer depends on who happens to be writing it up that day