Open to opportunities

anmol gautam

@anmolgautam

Message

Lead Applied Scientist building production-grade AI for enterprise search and agents.

India

Message

What I'm looking for

I’m looking to join a team that builds production-grade AI—multi-agent systems, RAG, and Text-to-SQL—then measures success via latency/cost and real user outcomes. I want to keep bridging research to deployment with strong engineering discipline.

I’m a Lead Applied Scientist with 4+ years of experience taking complex LLM and multi-agent systems from research prototypes to production. I focus on shipping real, measurable capabilities—enterprise search, Text-to-SQL, and autonomous agent workflows—designed for reliability, observability, and cost-aware performance.

At Veagle, I architected Neutrino, a multi-agent AI platform for enterprise search and Text-to-SQL, including human-in-the-loop execution and multi-LLM orchestration. I fine-tune domain-specific LLMs using LoRA/DoRA/PEFT with alignment approaches (DPO/GRPO), and I optimize latency and compute using benchmarking, pruning, and quantization with vLLM and SGLang—deploying across five major ISV partners.

Previously at 8bit.ai/SuperAGI, I built SuperCoder2.0, an autonomous code navigation and issue-resolution system that reached 33% on SWE-Bench-Lite using custom RAG and code generation. I also developed an instruction-tuned SAM-7B model and contributed an open-source runtime (Dendrux) for real-world agents with tool calling, persistence, and FastAPI/SSE hosting, alongside document AI and RAG-based systems that improved extraction and vision performance.

Experience

Work history, roles, and key accomplishments

Current

Lead Applied Scientist

Current

8bit.ai

Oct 2024 - Present (1 year 8 months)

Architected Neutrino, a multi-agent AI platform for enterprise search and Text-to-SQL, built with FastAPI/SSE and multi-LLM orchestration; deployed across 5 major ISV partners. Fine-tuned domain LLMs with LoRA/DoRA/PEFT and alignment (DPO, GRPO) and delivered agentic ReAct workflows for partner-specific RAG and Text-to-SQL.

Multi Agent Systems RAG Text To SQL LoRAs DORA PEFT DPO fastAPI VLLM

Applied Scientist

SuperAGI

Nov 2023 - Oct 2024 (11 months)

Built Text-to-SQL and RAG-based conversational multi-agent systems for SuperSales. Developed SuperCoder2.0, achieving 33% on SWE-Bench-Lite using custom RAG and code generation, and built an instruction-tuned SAM-7B (Mistral-7B) with GPT-3.5-comparable performance.

RAG Text To SQL SWE Bench React Mistral Custom RAG Model Evaluation

Associate Consultant

Dendrux

Aug 2022 - Oct 2023 (1 year 2 months)

Built an open-source runtime for real-world agents, supporting tool calling, persistence, observability, and FastAPI/SSE hosting with client tool-bridge pause/resume execution. Developed document AI and information extraction pipelines using OCI Document Understanding and EasyOCR (improving NER/key-value extraction by 7%) and delivered RAG/vision QA with a face recognition pipeline improving perfo

Runtimes Calling Tools fastAPI SSE Observability OCI Document Understanding RAG Computer Vision Transfer Learning

Research Intern - Nvidia

NVIDIA

May 2021 - Apr 2022 (11 months)

Collaborated on NLP and computer vision systems using NVIDIA NeMo and Hugging Face, including English-to-Hindi machine translation, object detection, and image segmentation. Achieved SOTA results published in IEEE by improving UNet performance and receiving the Institute Best Master's Thesis Award.

NVIDIA NeMo Hugging Face Transfer Learning UNet Segmentation Object Detection Machine Translation PyTorch Computer Vision NLP U Net

Education

Degrees, certifications, and relevant coursework

National Institute of Technology Meghalaya

Master of Technology (M.Tech), Computer Science and Engineering

2020 - 2022

Grade: 10.0 CGPA

Activities and societies: Gold Medalist (Academics); Institute Best Master's Thesis Award (Region of Interest Segmentation in Biomedical Images).

M.Tech in Computer Science and Engineering at NIT Meghalaya, graduating with a CGPA of 10.0/10 and earning Gold Medalist recognition. Received the Institute Best Master's Thesis Award for work on region of interest segmentation in biomedical images.