Looking for a job

ajeya krishna

@ajeyakrishna

Message

Senior AI Engineer at BukuWarung who shipped a RAG agent handling 3,000+ daily conversations with 80% containment.

Indonesia

Message

At BukuWarung, I shipped Sekar, a customer-support RAG agent handling 3,000+ conversations a day with 80% containment, 75% answer acceptance, and 80% CSAT. I also developed Cortex, a company-wide retrieval platform for 150+ employees, with permission-aware graph-and-vector retrieval.

At BukuWarung, I built retrieval evaluation and governance systems, including an org-wide policy gateway for PII redaction and audit trails. I also worked on retrieval-grounded voice agents and a multi-agent Sales Activity Intelligence platform.

At SaaS Labs, I fine-tuned and deployed models for meeting intelligence, email priority classification, and real-time spam classification. I owned their lifecycle from labeling and baseline modelling through production monitoring and retraining, and also built an AI Voice Agent that generated $1.5M in revenue.

Experience

Work history, roles, and key accomplishments

Current

Senior AI Engineer

Current

BukuWarung

Jul 2025 - Present (1 year 3 months)

Owns the retrieval stack end-to-end, including hybrid vector search, permission-aware retrieval, and indexing pipelines. Shipped a RAG agent resolving 80% of 3,000+ daily conversations and a knowledge platform for 150+ employees.

Python fastAPI LangGraph Lang Chain Qdrant Zilliz ArangoDB MCP Ranx Ragas Promptfoo PostgreSQL Redis Docker AWS GCP

Machine Learning Engineer-II

SaaS Labs

Jan 2022 - Jul 2025 (3 years 6 months)

Built multi-vector RAG pipelines and NLP extraction services across FastAPI and Docker microservices. Fine-tuned transformer models for tone scoring, email triage, meeting intelligence, and spam classification, and owned the end-to-end ML lifecycle from labeling to deployment and monitoring.

Python PyTorch Hugging Face Transformers BERT RoBERTa LLaMA 3 Lang Chain Weaviate Qdrant fastAPI Docker Kubernetes GCP

Education

Degrees, certifications, and relevant coursework

Indian Institute of Technology (BHU), Varanasi

B.Tech + M.Tech, Engineering Physics

Activities and societies: Published 2 papers on stochastic particle systems; applied transfer learning (ResNet50, EfficientNet) to medical image classification.

Completed the Integrated Dual Degree (IDD) program in Engineering Physics, with research on stochastic particle systems and applied transfer learning to medical image classification.