Acquired by Wisprflow
• Lead engineering effort and co-lead product. Lead a team that has designed and built multiple agentic systems including an email writing agent (Yapify - https://www.yapify.ai/), an agentic search system over CRMs (Atrium - https://www.atriumlabs.dev/), and an open source RAG data pipeline (Vectorflow - https://github.com/dgarnitz/vectorflow). Built fullstack applications using React, Next JS, FastApi, Postgres, Terraform, AWS.
• Raised $1M
Technical achievements include:
• Developed a hybrid retrieval architecture combining sparse and dense embeddings, dynamic prompt injection, HyDE augmentation, use-case-specific routing, hierarchical summarization, citation verification, data enrichment, LLM-powered ranking functions, and entity extraction to achieve high recall and precise filtering
• Engineered large-scale ingestion pipelines (AWS ECS, SQS, Postgres, Celery) capable of processing 20+ years of multimodal data including PDFs, emails, and CRM records, resulting in indexes with 1M+ embeddings
• Developed benchmarking & automated evaluation pipelines to tune retrieval & chat quality
• Developed an autonomous “plan-and-execute” agent for workflow automation across Slack, Notion, WhatsApp, Email, and Calendar; initially built with MCP and later refactored into modular sub-agents
