Talent acquisition leader building AI tools to fix hiring from the inside.
19 years of full-cycle recruiting across India, the US, Canada, and the UK. I've sat on the side of the table where the system breaks: resumes filtered before a human ever reads them, ATS configs that hurt candidates and companies alike, job descriptions that don't say what the team actually needs. So I do two things about it — I build tools that fix the system, and I teach job seekers how to beat it in the meantime.
Author, Rumblings of a Recruiter. Based in Toronto. LinkedIn Top Voice (2024). Top Recruiting Voice and Top Interviewing Voice (2023).
Three beliefs run through everything I publish, coach, and build:
-
Go deeper, not broader. The candidate who researches a company's actual problem statement beats the one who sends 200 applications. The recruiter who hunts outside conventional pools beats the one keyword-matching pedigrees. Depth wins where volume fails.
-
AI accelerates. People matter. "AI may help me accelerate. But it's people who help me feel." I use AI everywhere in my practice — pipelines, intelligence engines, agents — but always as a multiplier for human judgment and empathy, never a substitute for it.
-
Personal experience → framework → community. Every framework I publish started as a lived moment. A 2016 bet on a failed founder became my unconventional-sourcing thesis (he was CTO within a year). A college mentorship became a community initiative. I build outward from real stories, then give the frameworks away.
| Project | What it does |
|---|---|
| EchoLayer | Long-term memory for Claude agents — captures your work, distills lessons, and briefs Claude automatically at the start of every session. Local-first (nothing leaves your machine), benchmarked at 100% recall on 65% fewer tokens, 127 tests |
| career-ops | AI-powered job search system built on Claude Code — 14 skill modes, JD scoring, Go dashboard, PDF generation, batch processing |
| NewspaperEngine | News intelligence engine ingesting 23 sources (Reuters, Bloomberg, FT, arXiv, HN) and surfacing hidden connections via entity extraction and graph algorithms |
| claude-wiki-system | Automatic Obsidian-compatible wiki for all your Claude activity — daily logger, wiki viewer, backup/restore. Persistent memory for your AI workflow |
| anti_hallucination | Claude skill enforcing factual accuracy, uncertainty calibration, and source discipline — separating known facts from inference and speculation |
| learningskills | Claude skill for structuring AI prompts using 4MAT, Dehaene's Four Pillars of Learning, and Make It Stick principles |
| JobAutoFill | Saves job application fields once, reuses them everywhere — less time on forms, less job-search stress |
| ZenCycle | Wellness app helping professionals monitor focus time and remember to take breaks |
Talent acquisition (19 years)
- Full-cycle technical recruiting, TA operations, employer branding — North America and EMEA
- Talent intelligence at scale: longitudinal tracking of 4,800+ Ontario SaaS companies and 430K tech professionals using LSTM forecasting, graph neural networks, BERT topic modeling, and survival analysis
- ATS dysfunction diagnosis, AI recruiting maturity assessment, senior technical pipeline construction
- Recruitment-as-a-Service for Canadian SaaS and tech companies (50–500 employees)
AI engineering
- Claude skills, agents, and MCP servers — a full production skill ecosystem for TA and career work
- Python, Node.js, DuckDB, SQLite, document pipelines, automation infrastructure
- AI governance and the intersection of AI capability with human work
- Why your resume gets filtered before a human reads it (and how to fix it)
- What engineering managers actually mean when they write a job description
- How ATS systems rank candidates — and what breaks the algorithm
- Why "culture fit" rejections happen and how to reframe your story
- How recruiters, hiring managers, and candidates each optimize for different things — and how to use that game theory to your advantage
| 900+ | job seekers coached on Topmate — 4.9/5 across 135+ reviews, Top 50 Career Strategist in North America |
| 154,000+ | members in the Get Hired in Canada (GHIC) community |
| 97% | requisition fill rate; 3:1 interview-to-hire ratio |
| 28% | average reduction in time-to-hire delivered at scale |
| 19 yrs | full-cycle recruiting across India, US, Canada, UK |
- Substack — Career Clarity with Sajithkumar: career strategy, hiring market data, and the Rumblings of a Recruiter series
- Book — Rumblings of a Recruiter, a completed non-fiction manuscript on two decades inside the hiring system
- Research — practitioner reports on AI's state of play, the 2026 hiring market, and LinkedIn's algorithm shift
- Career coaching → topmate.io/sajithkumars
- LinkedIn → linkedin.com/in/sajithkumars
- Location → Toronto, ON