Uma candidatura completa num minuto
Currículo e carta de apresentação personalizados, prontos a enviar.
Execo in Noida is seeking a senior technical architect to design and build our internal agentic harness. You will lead delivery of initial agentic applications, including contract analysis and SAP/Ariba integrations, while ensuring security and governance.
The role focuses on building a scalable, model-agnostic framework and mentoring engineers in agent development. The ideal candidate has 10+ years in software, 4+ years in architecture, and hands-on experience with production-grade LLM systems
Noida, Pune Or Mumbai, India | Posted on 11/18/2026
Execo helps the world’s leading companies solve complex problems in areas like legal, go-to-market, and data annotation by combining world-class talent with the power of AI. Our teams work across seven countries to bring real innovation to life, turning the promise of AI into practical, high-impact solutions for clients every day.
At Execo, you’ll join a global community built on curiosity, ownership, and growth. We are a team that moves fast, learns continuously, and celebrates our progress together.
Join us to build what’s next in the AI-powered world of work and grow your career while helping businesses everywhere work smarter.
We are looking for a technical architect todesign and build our internal agentic harness: the shared framework that letsAI agents plan, call tools, read company data, act in our systems and beevaluated and governed safely. On top of that framework, the architect willlead delivery of the first internal agentic applications (for example, contractanalysis, integration support for SAP S/4HANA and Ariba, pre-sales and solutiondesign, and support triage).
The harness is the product. Individualagents should be quick to build because the hard parts are solved once: modelaccess, tool and data connectors, memory, permissions, evaluation,observability and human approval.
Area
Whatthe architect must be able to do
Priority
Design agent loops (plan, act, observe),multi-agent and sub-agent patterns, workflows vs. autonomous agents, and knowwhen an agent is the wrong tool
Must have
Prompt and system-prompt design,structured outputs, tool/function calling, context-window management,caching, streaming, model routing across providers
Must have
Tools and integration
Build tool schemas and MCP servers; wrapREST, OData and SOAP APIs safely; handle auth (OAuth 2.0, service accounts)and rate limits
Must have
Retrieval and data
RAG design: chunking, embeddings, hybridsearch, reranking, vector stores; document parsing (PDF, Word, scannedcontracts); permission-aware retrieval
Must have
Evaluation
Build eval sets from real tasks,LLM-as-judge and rule-based graders, regression suites in CI, A/B and onlinemetrics
Must have
Safety and security
Prompt-injection and data-exfiltrationdefenses, least-privilege tools, sandboxed code execution, approval gates forside-effecting actions, PII handling
Must have
Observability
Tracing of every model and tool call,cost and latency dashboards, replay of failed runs, feedback capture
Must have
Production Python and/or TypeScript;async and event-driven design; queues, workflow engines, idempotency andretries
Must have
Cloud and platform
Containers, Kubernetes or serverless,infrastructure as code, secrets management, CI/CD on AWS, Azure or GCP
Must have
Memory and state
Short- and long-term memory design,durable state for long-running agents, checkpointing and resumption
Agent frameworks
Hands-on with at least one framework(e.g. LangGraph, Claude Agent SDK, OpenAI Agents SDK, Semantic Kernel) andthe judgment to build thin abstractions instead of locking in
Cost management
Token budgeting, model tiering, caching,batch processing, unit economics per agent
Fine-tuning and model ops
When to fine-tune vs. prompt vs.retrieve; open-weight model hosting
Nice to have
Experience requirements