Agent Framework Lead

Experience
: years

Salary
: INR / year (based on experience)

Expected Notice Period
: 7 Days

Shift
: (GMT+05:30) Asia/Kolkata (IST)

Opportunity Type
: Office (Bengaluru)

Placement Type
: Full Time Permanent position(Payroll and Compliance to be managed by: Roots Health)

**(*Note: This is a requirement for one of Uplers' client - Maintaining patient trust and legacy)

**What do you need for this opportunity?

Must have skills required:
Multi-agent systems / distributed AI orchestration, FastAPI, Kafka, Python

Maintaining patient trust and legacy is Looking for:****
The Agent Framework Lead is one of the nine founding engineers responsible for designing the core intelligence layer - the framework that allows AI agents to perceive, reason, act, and self-evaluate.

You will architect the system that powers all Independence OS agents: onboarding agents, RCM agents, scribe agents, and future autonomous healthcare workflows.

This is not a "feature" role. This is a foundational systems role — you'll build the rails for how all agents think, learn, and communicate. You'll work at the intersection of LLM orchestration, stateful memory, tool-use, and evaluation harnesses — creating the backbone that will define how every agent behaves across healthcare, finance, and intelligence domains.

Your work will be open-sourced internally as the Mercor Agentic Framework — powering the world's first "AI- Native Enterprise Stack."

What You'll Build -
You will lead the design, architecture, and implementation of the Agent Runtime Layer and its supporting libraries.

• Your framework will unify how we:
• Orchestrate intelligence:
• Manage multi-agent workflows with Sierra-style orchestration (delegation, retries, idempotency).
• Define interaction protocols between task, reasoning, and evaluator agents.
• Establish an agent ontology:
• Design schema for agent types, roles, and state transitions.
• Enable knowledge continuity between specialized agents (handover protocol).
• Enable adaptive learning:
• Build eval feedback loops — every action logged, scored, and used for improvement.
• Integrate with continuous deployment (CI/CD + Evals) to measure agent performance in production.
• Power real-world execution:
• Tool calling, sandboxed code execution, and secure API access.
• Built-in observability, privacy hooks, and human-in-the-loop guardrails.
• The outcome of your work: agents that are self-evaluating, composable, and deployable in regulated industries.

Your Responsibilities

• Architect the Agent Framework Core, defining runtime abstractions for LLM-driven decision systems.
• Design the evaluation feedback layer, embedding "eval-first" principles into every build.
• Collaborate with the Evals Lead to define scoring, metrics, and success taxonomies.
• Partner with Full Stack and Platform Engineers to productionize the agent SDK.
• Ensure compliance-by-design: privacy, data redaction, audit trails, and safe orchestration.
• Maintain high reliability and low latency under concurrent load.
• Mentor early hires, codify best practices, and build internal playbooks.
• Lead demos and publish technical write-ups as part of our Frontier Learning Culture.

Your Tech Stack
Preferred Core Expertise:

• Python (FastAPI, LangChain, or custom orchestration layers)
• LLM APIs: OpenAI Function Calling, Anthropic, Gemini, or open-source LLMs (Mistral, Llama 3)
• Message queues & orchestration: Celery, Kafka, or custom async runtimes
• Vector Databases: Pinecone, Weaviate, Chroma
• Observability: Prometheus, Sentry, OpenTelemetry
• Cloud: AWS / GCP / Azure (Kubernetes, Docker)
• Testing/Evals: Harnesses for latency, cost, and accuracy metrics

Bonus Expertise:

• Multi-agent systems / distributed AI orchestration
• Reinforcement Learning from Human/AI Feedback (RLHF/RLAIF)
• Secure tool calling, retrieval augmentation (RAG), or ontology-based systems
• Prior work in Palantir Foundry, AutoGPT, or enterprise-grade LLM frameworks

What Success Looks Like
30 Days:

• Draft agent runtime spec and initial orchestration pipeline (state, retries, memory).
• Integrate eval logging into one internal agent (e.g., Meeting Pulse).

60 Days:

• Launch v1 of the Mercor Agent Runtime SDK.
• Connect agents with evals harness; measurable improvement in precision/latency.

90 Days:

• Deliver production-ready agent orchestration system.
• Demonstrate 2 real-world use cases with measurable evals (RCM, onboarding agent).
• Publish The Agent Framework Manifesto internally — documentation + codebase.

What We Value

• Systems thinking: You think in architecture diagrams and life cycles, not single scripts.
• Frontier hunger: You're obsessed with agent behavior, not just API responses.
• Builder DNA: You write, test, and ship — you don't wait for perfect conditions.
• Evals-first mindset: You measure success with precision, not perception.
• Ethical clarity: You design systems safe enough for healthcare-grade deployment.

You Might Thrive Here If

• You've shipped production systems with LLMs and orchestration.
• You see "multi-agent collaboration" as the next operating system paradigm.
• You prefer designing frameworks to fixing scripts.
• You believe India can produce the best AI-native engineers on Earth.
• You want your code to shape the rails of healthcare, not just apps.

Example Outcomes (Next 6 Months)

• Launch of Mercor Agent Runtime (internal SDK).
• Two agents (Pulse + Scribe) deployed in clinical settings.
• Evaluation dashboards show measurable lift in precision, recall, and latency.
• Publication of Agent Intelligence Architecture v1 — the blueprint for next-gen AI-native systems.

Compensation & Ownership

• This is a founding engineering role — with ownership, not employment.
• Competitive salary + equity in Independence OS / Mercor entity.
• Work directly with CTO and founding leadership.
• Build a framework that becomes the foundation of every AI-native enterprise.

How to apply for this opportunity?

• Step 1: Click On Apply And Register or Login on our portal.
• Step 2: Complete the Screening Form & Upload updated Resume
• Step 3: Increase your chances to get shortlisted & meet the client for the Interview

About Uplers:
Our goal is to make hiring reliable, simple, and fast. Our role will be to help all our talents find and apply for relevant contractual onsite opportunities and progress in their career. We will support any grievances or challenges you may face during the engagement.

(Note: There are many more opportunities apart from this on the portal. Depending on the assessments you clear, you can apply for them as well).

So, if you are ready for a new challenge, a great work environment, and an opportunity to take your career to the next level, don't hesitate to apply today. We are waiting for you