Ph.D. in CS / Sr. Staff Machine Learning Engineer @LinkedIn 🇺🇸 The opinions are my own. Husband, father of 2 boys
Saratoga, CA
Joined June 2010
- Tweets5.4K
- Following328
- Followers298
- Likes974
Pinned Tweet
We shipped semantic search for LinkedIn Hiring Assistant over 1B+ member profiles: free-text qualifications, no relevance labels at launch, only InMail engagement as a proxy.
A few things surprised us. Thread 🧵
Full write-up: linkedin.com/blog/engineerin…
Classic recruiter search trade-off: faceted search had coverage but many bad matches; boolean had precision but ~half of queries returned nothing.
EBR-powered semantic retrieval became the best single strategy — faceted-level liquidity with the best pre-ranking relevance.
Other lessons in the post: contrastive post-training alignment beat model size for embedding FT; offline judge replay predicted online A/B direction; relevance supervision at retrieval + engagement at ranking are complementary.
linkedin.com/blog/engineerin…
While chatting with a friend recently, I finally managed to pinpoint the one real edge humans still have over AI: taking full responsibility for decision-making. Everything else feels like the moat has already shrunk to almost nothing
Built my first AI-generated video technology #CareerReel end-to-end at an internal LinkedIn hackathon. Treated it like Lean Startup: ship fast, validate with real viewers, iterate hard. All learnings in 3‑min recap video 👇 #GenAI #Higgsfield #VideoResume t.me/ai_review/164
"High-Dimensional Probability" -- the Second Edition by Roman Vershynin is out! youtu.be/mSETfacxn80?si=J_K6… I am a huge fan of the first ed.
All set for @aclmeeting Vienna! #ACL2025 If you’re into industry-scale embedding-based retrieval, LLMs for search and recsys, LLMs-as-judge, or @LinkedIn’s semantic search and agentic experience for Recruiter Search, let’s connect
GPU retrieval can rival HNSW / IVFPQ on both latency and recall. @LinkedIn's LiNR (Feed OON recsys) and SJS and @Meta's MoL already deploy exhaustive k-NN at production scale on A100/H100 nodes. See my full post for the technical details ⬇️ t.me/ai_review/162 #VectorSearch
My two cents to the famous discussion: LLMs imitate reasoning (it’s how they are trained after all), but it works for practical use cases, as we saw @LinkedIn while working with job qualifications, often qualitatively better and surely faster than human. Fake it until you make it
I just realized that @AnthropicAI’s Claude is named after Claude Shannon. Beautiful
Our paper “A Scalable and Efficient Signal Integration System for Job Matching” on combining LLMs and GNNs for job recommendations @LinkedIn was accepted for #KDD2025 kdd2025.kdd.org/applied-data… see you in Toronto
Still true in 2025
What is missing:
Pros:
- betting on open source, which will beat any private company in the long run;
- wider applicability;
- amortized costs;
- building in-house expertise for p17n and embracing latest advances.
Cons:
- Tempting shortcut to reduce go-to-market time. #LLM
This is our team’s work on #LLM productionization for #RecSys from a year ago. Since September 2024, it has powered the MME in job recommendations and search @LinkedIn
linkedin.com/blog/engineerin…
While I don't mean to sound overly pessimistic, traditional chemical-based rocket engines are not sufficient for the demands of deep space exploration.Meanwhile, the rare and technologically complex enriched uranium-235 is being burned at low efficiencies in nuclear power plants
