Sarosh H.

Most AI projects fail in production — not because the model is wrong, but because the surrounding system isn't built to handle it. I fix that. I build and ship production-grade AI systems: LLM-powered agents, RAG pipelines, and full-stack platforms that enterprises actually use. My background is 20+ years of corporate engineering: healthcare, legal, fintech, SaaS — where systems have to be secure, auditable, and reliable from day one, not eventually. I'm one of Upwork's Expert-Vetted engineers, a designation earned through manual technical review that places me in the top 1% of the platform. I lead a tight senior team across AI/ML, full-stack, and cloud, which means you get my direct involvement on architecture and critical decisions, with the capacity to move faster than a solo hire ever could. 𝐖𝐡𝐚𝐭 𝐈 𝐛𝐮𝐢𝐥𝐝: → 𝐀𝐈 𝐀𝐠𝐞𝐧𝐭𝐬 & 𝐑𝐀𝐆 𝐒𝐲𝐬𝐭𝐞𝐦𝐬 — LLM integrations with retrieval, evaluation harnesses, prompt architecture, guardrails, and agent workflows built on Azure OpenAI, Anthropic, and OpenAI → 𝐅𝐮𝐥𝐥-𝐒𝐭𝐚𝐜𝐤 𝐏𝐥𝐚𝐭𝐟𝐨𝐫𝐦𝐬 — SaaS products, marketplaces, and internal tools in Next.js, Node.js, React Native, and Python — from schema to deployment → 𝐃𝐚𝐭𝐚 & 𝐁𝐚𝐜𝐤𝐞𝐧𝐝 𝐈𝐧𝐟𝐫𝐚𝐬𝐭𝐫𝐮𝐜𝐭𝐮𝐫𝐞 — APIs, pipelines, event systems, Postgres/MongoDB/Redis, Elasticsearch, Kafka — designed for performance and cost governance → 𝐂𝐥𝐨𝐮𝐝 & 𝐒𝐞𝐜𝐮𝐫𝐢𝐭𝐲 — AWS, Azure, GCP; Docker/Kubernetes; Terraform; SOC 2, HIPAA, and GDPR-aligned architecture with SSO, RBAC, and audit logging 𝐂𝐥𝐢𝐞𝐧𝐭𝐬 𝐜𝐨𝐦𝐞 𝐛𝐚𝐜𝐤 𝐛𝐞𝐜𝐚𝐮𝐬𝐞 𝐭𝐡𝐞 𝐰𝐨𝐫𝐤 𝐡𝐨𝐥𝐝𝐬 𝐮𝐩. Clean architecture, measurable outcomes, full documentation, and handovers that don't leave your next engineer guessing. If you're building something that needs to work in the real world — not just demo — let's talk.