When legal intake happens everywhere at once
Building a no-code Slack workflow to triage questions, reduce noise, and stay in the loop.
đ Hey there, Iâm Hadassah. Each week, I unpack how in-house legal teams use AI to enable the business, protect against risk, and free up time for the work they enjoy mostâwhat works, what doesnât, and the quick wins that make all the difference.
Before we dive in, a quick note: this is just one example of a legal team solving an operational bottleneck. There are plenty of ways to approach these kinds of problems, and the right solution will always depend on your specific needs and context. My goal is to give you some food for thought as you define what that solution should be.
Problem
Weâre joined by the GC at a software scale-up who was drowning in questions long before they ever reached the âlegal issueâ stage. On any given day, 20â30 Slack pings arrived across open channels, mixed with comments buried in Google Docs, Jira tickets appearing out of nowhere, spontaneous Trello board mentions, email follow-ups, and the occasional shoulder tap in the hallway.
The problem wasnât just volume, it was visibility. Legal advice shared in one channel would reappear elsewhere, distorted or oversimplified: âLegal said we canât do this.â Sometimes that was true; often it wasnât. People invoked âLegalâ to shut down debates, or repeated guidance out of context, unintentionally creating confusion the GC then had to unwind. Meanwhile, collaborative tools introduced their own friction. Google Docs created version-control nightmaresâcomments disappeared, drafts were overwritten, and redlines were lost when someone made a copy and deleted the original.
The result was a level of noise that made it nearly impossible to prioritise work, maintain consistency, or understand how the legal team was being represented across the company. The GC was carrying what had effectively become a 150-person intake system without any structure to support it.
Solution
The GC needed a way to bring order to the chaos without adding headcount, budget, or engineering support. The answer came in the form of a no-code workflow built entirely from tools the business was already using:
Stitched together, these tools were used to centralise incoming requests, summarise conversations, and automate first-pass answers while keeping a human firmly in the loop.
The goal wasnât to replace legal judgment. It was to create visibility: a single place where every legal reference, question, and misunderstanding surfaced. From there, the system could filter and prioritise the signal, and only then automate the parts that drained the most time.
Letâs dig a little deeper: Zapier became the engine behind the workflow. It monitored Slack channels for mentions of the GC or the legal team and quietly captured each oneâquestions, references to prior guidance, even informal nods to âLegal wonât like this.â These messages were funnelled into a single Google Doc each day, replacing hours of manual thread-checking with one consolidated feed.
From there, the AI layer took over. At the end of each day, the GC ran the document through their enterprise ChatGPT environment to summarise, highlight sentiment, and stack-rank issues by urgency. What once required scrolling through 20+ pages of Slack history now surfaced in a short, structured brief the GC could review in minutes.
For direct questions, the solution went a step further. ChatGPT generated a first-pass answer, which was sentâthrough Zapierâto a private Slack channel visible only to the GC. From this channel, the GC could approve the answer with a simple đ emoji. Zapier then delivered the vetted message back to the original requester, written as if the GC had typed it personally. Nothing went out automatically; every answer carried the GCâs voice and oversight.
And the foresight didnât stop there. The GC planned a next layer: a simple knowledge repository that would store validated answers and reuse them before generating anything new. Over time, this would turn the system into a lightweight internal legal OS built entirely on a no-code infrastructure.
Results
20â30 daily Slack pings were consolidated into a single triage view, cutting manual monitoring time dramatically.
Daily AI summaries provided instant visibility into developing issues, risks, and internal sentiment toward the legal team.
The GC could correct misunderstandings quickly, before they spiralled into team-wide assumptions or became talking points in leadership meetings.
First-pass AI-powered answers, paired with a simple thumbs-up approval, gave employees consistent responses in a fraction of the time.
Even without a sophisticated intake system, the GC got access to valuable data: how often the legal advice was invoked, how many questions came in, and which topics surfaced repeatedly.
Process
The GC didnât begin by building a full system, but rather by identifying a single pain point: Slack was where the noise was loudest. So the first step was simply monitoring mentions and collecting them in one place. This alone created an immediate informational win: the GC could finally see the full landscape instead of scanning channels reactively.
From there, the solution evolved gradually. Once the GC saw that centralisation worked, they layered in AI summarisation to reduce review time even further. Then came stack-ranking to highlight the top five items. Only after that did they introduce AI-generated first-pass answersâand even then, only with a mandatory human review step to preserve accuracy and trust.
Security and privacy were handled upfront. The GC ensured that the model didnât retain or train on company data and that everything passed through an internal approval step. Because the company encouraged employees to build their own internal tools, stakeholder friction was minimal; the GC obtained Slack access approvals where needed, and the rest fell naturally into place.
For end users, nothing changed. Responses still came from the GCâs account, in the their tone, through the channel they already used. This sidestepped the usual adoption hurdles entirelyâno new UI, no training sessions, no behaviour change. The experience felt personal, even though the system behind it was doing the heavy lifting.
Quick Wins
Not surprisingly, this solution wasnât built in one push; it was shaped over months through repeated small improvements, each one nudging the system closer to a scalable intake and triage layer. The quick wins this GC relied on to build moment and keep the project moving were:
Using accessible, visual tools. Zapier was easy to use and light-touch, making it approachable even for someone without extensive coding experience. Its visual, modular nature allowed for setting up "if this happens, then that" logic effectively. This enabled the initial steps of gathering data.
Human-approved automated responses. A simple đ triggered fast, consistent answers while keeping legal judgment firmly in the loop. This served as a check and balance against AI hallucinations, ensured the accuracy and ethical compliance of the advice given, and allowed the GC to make the final output their own.
An iterative, step-by-step building approach. The solution was built incrementally, starting with small, achievable improvements. The first was simply gathering scattered Slack communications in one channel. Subsequent steps included summarisation, ranking importance, drafting individual answers, and finally automating the response process. This iterative approach allowed for learning and refinement along the way without needing to build a perfect solution upfront.
Now itâs your turn. If your team is dealing with something similar, I hope this story sparks a few practical ideas you can put to work.
And⊠if youâve been through something similarâor solved a different operational challenge altogetherâIâd love to hear your story and spotlight your win.