Skip to content

About

I'm Dhruv Makwana — an AI Engineer and Deep Learning Researcher with 5+ years turning research into production AI systems. I work at the intersection of applied research and deployed systems: publishing in peer-reviewed venues while shipping models that run in production at scale. This site is where I write up that work — LLM agents, tool use, evaluation, and production engineering, in the depth I'd want if I were reading it myself.

What I work on

At Centific, I design and deploy AI systems across LLMs, computer vision, and multimodal learning, including:

  • Redesigned a law enforcement evidence-review pipeline (video, audio, documents), cutting officer review time ~85% by combining deterministic techniques with narrowly-scoped reasoning-model calls, deployed on-premise for data privacy.
  • Built Spanglish (Spanish-English code-switched) ASR for law enforcement, reaching 0.1 WER through synthetic data generation and domain-specific fine-tuning.
  • Designed SurgQ-VLM, a two-stage SFT → GRPO surgical video QA model with a novel dual-stream compressor, evaluated on a sealed 21,840-question benchmark.
  • Built a real-time multi-camera security analytics pipeline combining rule-based alerting with VLM judgment, reaching 89.5% correct aggression/weapon alerts in live evaluation.

Before Centific, at Ignitarium and Qualitas Technologies, I worked on 6DoF pose estimation for robotics, edge-device classification models, fault diagnosis from sensor data, and defect detection — optimizing for cloud, edge, and embedded deployment with TensorRT, TFLite, and mixed precision.

Research

Alongside industry work, I've been an active researcher through IIT Roorkee's CANDLE Lab, with peer-reviewed publications in venues including WACV, ICDAR, ISQED, and Elsevier, spanning low-light image enhancement, scene text understanding, PCB defect classification, and nuclei instance segmentation.

I'm looking to go deeper into research-driven roles — Applied Scientist / Research Scientist positions — where I can push on agentic systems, LLM alignment (RLHF/GRPO), and multimodal reasoning, while staying close to real deployment constraints.

About this site

A companion to RAG, Deep Dive — same standard: real, tested code where a technique can be demoed, current research over recycled tutorial content, and an honest report of what actually happened when the code ran, including the surprising and negative results. Every fast-moving claim (a library version, a protocol spec, a benchmark score) is stamped with the date it was checked, because this field moves within months.

Hiring, collaboration, or just talking shop

If you're hiring for an Applied Scientist / Research Scientist role, want to discuss a collaboration, or just want to talk about agents, RAG, LLM alignment, or multimodal systems — reach out. I read everything that comes in.