Case Study: AI implementation for a Singapore-based HR consultancy.
How we helped a boutique HR firm save 35+ hours per week by integrating AI into their daily operations.
Key results.
35+ hours saved per week. 3x faster candidate shortlisting. $6,000+ monthly cost reduction.
Client overview.
A boutique HR consultancy based in Singapore specialising in recruitment and talent advisory. A team of 15-20 serving clients across the region.
The challenge.
The firm had built a strong reputation through good work and word of mouth. But as the client base grew, the cracks in their operations became harder to ignore.
Candidate shortlisting was eating up the week. Each search required manually reviewing dozens of resumes against a job brief and writing up summaries. 8-10 hours per search before a single interview got scheduled.
Client reporting was inconsistent. Every client expected regular updates in a different format. The team spent hours each week writing what were essentially the same status updates with different names.
Proposals took too long. 2-3 hours to tailor each one, even though most of the content was the same every time.
Internal knowledge was scattered. Candidate notes, market insights, client preferences. All stuck in individual inboxes and spreadsheets. When someone left, their knowledge left with them.
The team had tried using ChatGPT a few months earlier. A couple of people used it for a week or two, mostly for rewriting emails. It never became part of how the firm actually worked.
Our approach.
A 5-week AI implementation engagement focused on the workflows costing the most time relative to the value they added.
-
We sat with the team. Shadowed the recruiters, watched how reports got put together, and mapped the full lifecycle of a typical engagement from pitch to placement. Identified 11 recurring workflows and prioritised four based on time consumed, how repetitive they were, and how much they affected client experience.
-
Designed solutions using off-the-shelf tools the team could pick up without any technical background. Claude for screening, drafting, and knowledge management. Make for connecting their existing systems. Granola for meeting capture. Set up a shared AI workspace calibrated to the firm's tone, document formats, and client expectations.
-
Candidate shortlisting: recruiters paste a job brief and resumes into Claude. It scores candidates against the brief, flags gaps, and produces a formatted shortlist ready to send. 8-10 hours per search down to 2-3.
Client reporting: prompt templates for each client's preferred format, connected to an automation that pre-fills data from their system. Recruiters just add commentary.
Proposals: Claude generates a tailored proposal from a few inputs. Gamma handles the slide version. 2-3 hours down to 30 minutes.
Knowledge management: interview and briefing notes captured through Granola, summarised by Claude, stored in a shared searchable workspace. First time the firm had institutional knowledge that wasn't trapped in someone's inbox.
-
Individual hands-on sessions with every team member using their actual workload. Not slides, not demos. We sat with each person and had them do it themselves. Trained one AI champion to maintain and extend the workflows going forward. Delivered a full playbook with every prompt, automation, and modification guide.
Results after 90 days.
Candidate shortlisting: 8-10 hours to 2-3 hours per search
Client reports: 45 minutes to 10 minutes each
Proposals: 2-3 hours to 30 minutes
Knowledge management: from scattered to searchable
Meeting notes: from inconsistent to structured.
35+ hours saved per week. Roughly $6,000 in monthly savings. The firm took on 12 new client engagements the following quarter without adding headcount.
“We were spending way too much time on admin instead of actually talking to candidates and clients. Amplify understood how we work and set us up with tools the team actually uses every day. It changed how we run the business.”
— Founder, HR ConsultancyWhy it worked.
We watched how the team actually spent their time, found where the hours were going, and built around those specific problems.
We kept everything simple. Off-the-shelf tools, no custom software, nothing that needs a developer to maintain.
We trained people using their actual workload, not slides about AI. That's what makes adoption stick.
And we ensured they were fully independent. The firm has already built new workflows on their own without us. That's the whole point.
Book a free discovery call.
Find out exactly where AI can save your team the most time and costs. 30 minutes, no commitment, and you'll walk away with actionable ideas whether we work together or not.
