Description:
About Pravar AI
Pravar is an outcomes-led AI transformation platform — helping enterprises close the gap between AI ambition and AI maturity through comprehensive Agentic OS platform, embedded domain teams, and purpose-built AI solutions.
Pravar is a well funded startup, with early traction and proven PMF transforming some of the biggest companies in the industry, one agent at a time.
Built by people who have sat on both sides of the table.
Pravar was not designed from a whiteboard. It was designed from decades of navigating enterprise complexity, building global products and leading companies through transformations.
Ayon Banerjee
Co-founder & CEO
Ayon brings over 30 years of experience in technology go-to-market and P&L management. He has built and scaled enterprise technology businesses across global markets, worked with Fortune 500 clients, and operated at the intersection of strategy and execution throughout his career. At Pravar, he leads the company's commercial direction and the overall transformation thesis — ensuring that every engagement is tied to outcomes, not activity.
Malavika Lakireddy
Co-founder and CPDO
Malavika has 22+ years of global product leadership across some of the most demanding technology environments in the world. Her track record spans Uber, Zeotap, AT&T, and 24/7.ai — building and scaling enterprise products across mobility, data intelligence, telecommunications, and conversational AI. At Pravar, she leads product development and client delivery, bringing the rare combination of product rigor and deep AI-era fluency needed to build a platform at the frontier of enterprise AI.
Job Description
We are hiring a Founding Senior/ Lead AI Platform Engineer to help design and build and extend the core PravarAI Agent operating system. Pravar OS is the technical spine of the product. You will design and develop the entire Agent OS — orchestration layer, multi-agent routing, context optimization, model selection, Guardrails and the evaluation harness. You are not just building agents; you are building the platform that scales the agents
Responsibilities
- Architect multi-agent orchestration platform (tools, MCP, routing, context management, agent handoffs)
- Own model optimization strategy — best-fit LLM routing across providers
- Drive 40–60% token reduction via context compression and prompt caching
- Build and maintain the evals/observability harness; define pass/fail criteria per agent
- Integrate MCP for structured data access across CRM, holdings, market feeds
- Build guardrails framework — policy enforcement, hallucination detection, output validation
- Build platform designed for on-prem or private cloud deployment
- Act as forward deployed engineers from time to time and accelerate platform building with client collaboration
- Mentor engineering team; drive code reviews and architectural discussions
Requirements
- 5 - 10+ years in ML/AI engineering; 2+ in LLM/agent systems in production
- Deep experience with LangGraph, CrewAI, AutoGen, or equivalent orchestration frameworks
- Strong Python, FastAPI, async programming
- Hands-on with OpenAI, Anthropic, Gemini APIs; fine-tuning open source models experience a plus
- Experience with RAG architectures, vector databases (Pinecone, Weaviate)
- Understanding of eval frameworks: LangSmith, RAGAS, or custom harnesses
- Bonus: prior work in fintech, wealth management, or regulated industries
Perks
- Hybrid role - focus on productively working vs spending time in traffic
- Best machines - We believe in giving you the best tools needed
- Claude max plans + AI tools
- High autonomy - you are the owner and the artist
Why Pravar
- Build real products on a production-grade AI platform
- Work at the intersection of cutting edge AI platform builder and driving outcomes with clients
- High ownership, low bureaucracy, direct impact
- Learn fast, ship fast, and grow alongside a strong engineering team
- Be part of a team transforming companies and building AI native organizations