Sattyam Jain
Tech Lead · AI Platform Architect · Production Agentic Systems
Ahmedabad, India · Daily US Pacific overlap · Open to relocation and visa sponsorship
sattyamjain96@gmail.com · github.com/sattyamjjain · linkedin.com/in/sattyamjain · sattyamjjain.in · theproductionagent.substack.com
Summary
Tech Lead and platform architect at Attri.ai owning engineering across 12+ active enterprise GenAI engagements in regulated US verticals (Insurance, Healthcare, AmLaw 200 Legal, EdTech, Construction, Geospatial). Architect of Agentify— Attri's flagship multi-agent platform with 15+ production agents — and the multi-cloud on-prem deployment mode that unlocked the regulated-vertical customer book. Creator of pyAGI (acquired by AGI House, 2025) and the Capability Lease security pattern (short-lived, scope-narrowed agent credentials). Ship and maintain the agent-security OSS layer in active production use: agent-audit-kit (OWASP Agentic Top-10 SAST), agent-airlock (runtime firewall), mnemo (Rust MCP memory DB). Author of The Production Agent, serialized 2026.
Experience
Tech Lead — promoted Oct 2025 from Senior Software Engineer
Dec 2024 – Present
Attri.ai · Ahmedabad, India · Remote with daily US Pacific overlap
- Architected Agentify — Attri's flagship production multi-agent platform with 15+ specialized agents (Orchestrator, PRD, Architect, Designer, Coder, Diff Analyzer, Discovery) for end-to-end product workflows.
- Designed the Orchestrator agent: central state-machine brain coordinating multi-agent flows, tool calls, and human-in-the-loop checkpoints with full replay + audit.
- Built the multi-provider LLM routing layer across Anthropic Claude (Opus 4.7 / Sonnet 4.7 / Haiku 4.5), OpenAI GPT-5.5 / 5.5 Pro, Azure OpenAI, Google Gemini 3 Pro — with fallbacks, token budgeting, and rate limiting.
- Engineered the Cascading Router cost strategy — ~70% LLM cost reduction by routing simple queries to small/cheap models and complex tool-use to Opus 4.7.
- Implemented E2B micro-VM sandboxing for agent code execution; led MCP-style tool integration across 10+ services with standardized auth, contracts, and result normalization.
- Designed the multi-tenant SaaS architecture: workspace segmentation, RBAC, Stripe billing, usage-based pricing, zero-downtime blue-green deployments — running at 99.9% production uptime.
- Architected the multi-cloud on-prem fork of the flagship platform: Terraform-per-environment, adapter pattern (Keycloak / MinIO / Postgres / Azure Blob), jsonb metadata for client schema variance — one codebase, N client clouds; unblocked the regulated-vertical deployment mode.
- Customer-facing principal engineer across 12+ enterprise engagements in 6 regulated US verticals (Insurance, Healthcare, AmLaw 200 Legal, EdTech, Construction, Geospatial); own scoping, integration design, and post-deployment governance.
- Greenfield delivery of a regulated healthcare AI product (FastAPI + React, empty repo → v1 in 6 weeks); authored the pre-launch Infrastructure + Compliance Audit closing PHI / secrets / observability gaps before customer onboarding.
- Coined the Capability Lease security pattern — short-lived, scope-narrowed, revocable agent credentials; adopted across agent-airlock, mnemo, and the AAK-CAPLEASE rule family in agent-audit-kit.
- Built attri-dev-kit, the org-wide CI/CD platform serving 208 internal repos with AI-specific guardrails (hallucinated-import detection, AI-author signature escalation, unjustified-lint-disable hard-blocks, test-delta gate).
- Built eval harnesses + OpenTelemetry → Datadog / Sentry / PostHog observability for agent reliability, tool-call accuracy, and latency across 9 production AI projects (Tech Owner).
- Allocated 25+ engineers across 16 simultaneous customer projects via a custom Resource Allocation Planner; performance-reviewed a 22-person engineering org; authored the org-wide Engineering Process Alignment doc adopted across every active project.
Senior AI Engineer
Dec 2023 – Nov 2024
Zenarate · Gurugram, India · Hybrid
- Shipped LLM-powered conversational AI features across multiple model families (private + open-source) for the contact-center training simulation platform.
- Cut p95 response latency 40% (1.6s → 960ms) by separating intent classification from generation and warm-pooling inference workers.
- Built predictive analytics pipelines on Azure CLU + LUIS; reporting accuracy lifted 15 points, improving client decision-making speed.
- Re-architected the simulation backend with health-gated rollouts and per-service circuit breakers; deploy time 18 min → 4 min, unplanned downtime −30%, sustained ~3k concurrent training sessions.
- Designed REST + WebSocket APIs (FastAPI) for the real-time simulation layer; managed AWS infrastructure-as-code with high-availability rollouts.
- Drove architecture decisions on service boundaries, session-tracking data models, and AI-assisted feature integration that did not compromise platform reliability.
Team Lead — promoted May 2023 from Software Developer
Dec 2020 – Nov 2023
MyShubhLife (FinTech) · Bengaluru, India
- Led the team building Kautilya LMS — a SaaS course + certification platform serving 500+ partner users with self-service enrollment, content delivery, certification, and proctoring.
- Built Heimdall from scratch — an internal ERP control tool giving partner ops real-time visibility across 50+ integration endpoints; cut partner onboarding 50% (≈4 wks → ≈2 wks).
- Designed and shipped backend services for loan processing, customer onboarding, and financial analytics for the consumer-lending product (Python + Django/DRF + Celery + Postgres + Redis).
- Built a production NER + OCR KYC pipeline (Python + AWS) processing 10,000+ documents/day at 92% extraction accuracy; cut human review queue 68% and unblocked same-day loan disbursal.
- Designed APIs + architecture for loan lifecycle, credit-scoring integrations, and reporting; tuned database performance + monitoring for sustained throughput.
- Co-authored pyluca — open-source double-entry accounting library (60+ ★); mentored engineers, ran code reviews, facilitated design discussions, and bridged technical + non-technical stakeholders.
Projects
Selected open source and side projects
- pyAGI — Acquired by AGI House, 2025. Early Python framework for agentic loops (planning / memory / tool use). Built 2023; acquired by AGI House (San Francisco) in 2025 — founding-member-led acquisition.
- agent-audit-kit SAST scanner for AI agents. Full OWASP Agentic Top-10 (10/10) + MCP Top-10 (10/10) coverage; SARIF output; 11-framework compliance mapping (EU AI Act / SOC 2 / HIPAA / NIST AI RMF / India DPDP / ISO 42001). GitHub Marketplace-listed; Sigstore-signed releases; public 48h CVE-to-rule SLA.
- agent-airlock Runtime firewall for AI agents — ghost-arg stripping, PII masking, RBAC, E2B sandboxing, circuit breaker, cost tracking. Enforces Capability Lease policies in <10ms p99. 1,157 tests · 9 framework integrations.
- mnemo MCP-native embedded memory database for agents (Rust). REMEMBER / RECALL / FORGET / SHARE primitives; hybrid vector + full-text search; AES-256-GCM encryption; branching/replay. Python / TypeScript / Go SDKs.
- MannSetu Voice-first AI mental wellness platform for India (Hindi / English / Hinglish). 50+ active users · 10K+ cumulative conversations; eliminates 2–4 week therapy wait times.
- VAJRA Physical AI side project — autonomous find-and-fetch robot on Jetson Orin Nano + ROS 2 + NanoOWL. Capability Lease governance instincts applied to physical actions.
Skills
- Languages:
- Python, TypeScript, Rust, SQL, Bash, Go (SDK-level)
- LLMs and agents:
- Claude (Opus 4.7 / Sonnet 4.7 / Haiku 4.5), GPT-5.5 / 5.5 Pro, Gemini 3 Pro · LangGraph · Model Context Protocol (MCP) · Anthropic SDK · OpenAI SDK · RAG · multi-agent orchestration · custom eval harnesses · LLM-as-judge · AgentOps · cascading-router cost optimization
- Agent security:
- OWASP Agentic Top-10, MCP Top-10, capability-scoped credentials (Capability Lease), prompt-injection defense, agent supply-chain auditing, CVE disclosure, SARIF-based SAST for agents
- Backend and data:
- FastAPI, Django/DRF, PostgreSQL, pgvector, Redis, Celery, DuckDB, USearch, Tantivy, vLLM (inference serving)
- Cloud, platform, observability:
- Azure (primary), AWS, Docker, Kubernetes, Terraform, Keycloak, MinIO, E2B microVMs, GitHub Actions, OpenTelemetry, Datadog, SigNoz, Sentry
- Compliance and regulated deployment:
- EU AI Act, SOC 2, HIPAA, NIST AI RMF, India DPDP, ISO 42001, ABA Model Rule 1.6 + Formal Opinion 512 (legal AI), on-prem + air-gapped deployment patterns
- Physical AI:
- ROS 2 Humble, NVIDIA Jetson Orin Nano, NanoOWL (TensorRT-optimized OWL-ViT), CUDA, behavior trees
Education
- Master of Computer Applications · Madhav Institute of Technology and Science (MITS), Gwalior · 2017 – 2020 · CGPA 8.9/10
- B.Sc. (Computer Science) · Bundelkhand University, Jhansi · 2014 – 2017
Talks and Writing
- The Capability Lease · sattyamjjain.in/writing/capability-lease — Manifesto + wire format for the pattern; adopted in agent-airlock and mnemo.
- The Production Agent (book, serialized in public) · sattyamjjain.in/book — 5 chapters on orchestration, MCP security, supply-chain audit, agent-on-agent commerce.
- OWASP Agentic Top-10 2026 Benchmark · sattyamjjain.in/benchmark/owasp-agentic-2026 — 10 commercial agent platforms scored across 10 OWASP Agentic families.
- Speaker availability: agentic AI, LLM evals, agent security, MCP governance. Request a slot at sattyamjain96@gmail.com.