Founding Software Engineer

Shyva · Stealth · Remote (India)

Shyva is building the verified trust layer for global trade — an AI-native procurement intelligence platform, in stealth mode.

We're looking for a Founding Engineer who thrives in ambiguity, ships fast, and has a genuine obsession with large-scale data systems. You'll work directly with the founding team to build Shyva's core platform: supplier discovery agents, entity resolution pipelines, semantic search, and the procurement intelligence layer.

This is a 0-to-1 role — broad ownership, direct influence on architecture, and a front-row seat to enterprise AI in global trade.

Applied AI Engineering

Shipped production AI products end-to-end — concept, architecture, evaluation, deployment, and ongoing ownership

Built AI systems that operate over both structured and unstructured data: retrieval, extraction, reasoning, or workflow automation

Designed confidence-aware systems with human-in-the-loop review where the stakes demand it

Sharp judgment on where LLMs add value (reasoning, extraction) and where strict deterministic engines must take over (financial calculations, regulatory countdowns, and tariff math)

LLM and Agent Orchestration

Shipped multi-step agent workflows in production with modern orchestration frameworks

RAG pipelines with hybrid retrieval and reranking

Guardrail architecture: post-generation validation, uncertainty flagging, stale-data detection

Search and Retrieval Systems

Production experience with modern search systems including vector and hybrid retrieval, reranking, and relevance tuning

Experience with knowledge graphs, entity linking, or multi-hop retrieval

Strong instincts for retrieval quality, explainability, and trustworthiness

Document Intelligence

Production experience extracting structured information from messy unstructured documents

Experience with entity resolution, record linkage, or deduplication across noisy real-world data

Large-Scale Data Engineering

Production ETL/ELT pipelines at scale

Experience ingesting and normalizing heterogeneous commercial data feeds with proper provenance and freshness tracking

Data lineage and auditability: every output traceable to source, timestamp, and confidence level

Full-Stack Engineering

Python backend mastery, coupled with strong modern frontend skills (React/Next.js, Tailwind). You can translate high-fidelity UI/UX concepts into premium, responsive enterprise dashboards.

Cloud-native deployment on AWS or GCP, containerization, CI/CD

Engineering and Systems Ownership

Strong software engineering fundamentals; ships production systems end-to-end

Comfortable building APIs, workflows, integrations, and pragmatic product-facing systems

Experience designing secure, multi-tenant architectures. Understands data compartmentalization, RBAC (Role-Based Access Control), and the technical requirements for enterprise compliance (e.g., SOC2).

Owns systems in production — reliability, observability, and the operational calls that come with it

Startup Execution

Comfortable operating in ambiguity and moving quickly without fully defined specs

Strong ownership mindset with pragmatic decision-making

Willing to challenge assumptions and make architectural trade-offs

CS, Engineering, or equivalent technical background

6+ years of hands-on engineering experience; track record of shipping end-to-end products

At least one role where you built something significant without a platform team or DevOps support

Recent experience at an early-stage AI startup shipping LLM-native products

Supply chain, procurement, or trade finance domain knowledge

Background in domains where data accuracy has direct financial or compliance consequences

Experience with enterprise system connectors (SAP Ariba, Oracle, or similar)

Has built or contributed to open source projects in search, retrieval, or document AI

Competitive compensation plus meaningful founding-engineer equity — range discussed with finalists

Fortune 500 design partners already committed — you will build for real customers from day one

Full architectural ownership: you decide the stack, the data model, the trade-offs

Remote, India-based, with at least 4 hours of daily overlap with US Central time

We're not looking for engineers who implement tickets. We're looking for someone who:

Has shipped something significant end-to-end and can point to it

Reasons about data quality and auditability as a first-class concern, not an afterthought

Understands that in enterprise procurement, a wrong number has real financial consequences — and designs systems accordingly

Is comfortable making architectural decisions in ambiguity and living with them

Has opinions about how to build this and will push back when they disagree

Send your resume and a brief note answering:

The most technically complex data system you have built and what made it hard

An architectural decision you made with incomplete information and why

What draws you to a role where the hardest problems are data quality and trust, not model performance