Skip to content
Shikher Jha

Always current. Generated from the same data as the site.

Download PDF ↓

Shikher Jha

Founding Engineer · Samora (YC W26) · Founder, StoreOS · Incoming SDE, Amazon

jhashikher@gmail.com · linkedin.com/in/shikher-jha-1b9853292 · github.com/shikherjha · Delhi / Bengaluru, India

Product engineer who builds the orchestrators, control planes and state machines that let software keep working when the real world doesn’t. Go, Python, Temporal, Kubernetes, Postgres, LLM systems.

Experience

Samora AI (YC W26) | Founding Engineer | Bengaluru2026 – Present
  • Architect an omnichannel human-in-the-loop platform (voice, SMS, WhatsApp, email, web chat): operator assignment, escalation, takeover and handback.
  • Designed Temporal-based human-layer orchestration with idempotent transitions and reaper/reconciliation logic that recovers stale jobs, abandoned sessions and partial failures.
  • Built call forwarding, voicemail detection and retry/fallback handling; hardened the concurrency-bounded dispatcher across SIP, WebRTC and LiveKit voice pipelines.
  • Built a post-call LLM evaluator that cites transcript evidence for every answer and publishes idempotent, typed campaign outcomes.
  • Rolled out observability across Lambda, ECS and databases (Datadog, OpenTelemetry, CloudWatch, Prometheus, Grafana); one call ID joins logs and traces.
StoreOS | Founder | storeos.in2026 – Present
  • Building the agentic operating system for ecommerce: Go core API, Postgres row-level-security tenancy, typed ChangeSets (propose, preview, approve, apply, undo), idempotent mutations.
  • Built the Kubernetes control plane for per-store Medusa deployments (Kopf, now a Go controller); prototype ran 10+ parallel provisions at >99% success, zero orphaned resources.
  • Designed the agent plane on Temporal (durable per-conversation workflows, build-and-review loop); cart, checkout and payment paths are locked to agents.
Auric AI | Software Development Engineer Intern | RemoteMay – Jul 2026
  • Owned the geospatial annotation-to-model pipeline: FastAPI data services over TIFF imagery, CVAT annotations and model outputs; hybrid automated/manual QC engine; Rasterio tiling pipelines; QC-approved annotations promoted into CV training sets.
Tealbox Digital | Backend & Automation Engineer Intern | RemoteMay – Sep 2025
  • Architected a production FastAPI backend (REST, webhooks, Redis async pipelines, PostgreSQL, Docker on Railway/GCP); cut manual toil by 90%.
  • Built Slack-native AI automations with n8n, LangGraph, Playwright and vector retrieval; cut analyst effort by 80%.

Projects

Signal (Go, FastAPI, React/TypeScript, pgvector, Redis, RabbitMQ)getsig.in
  • Turns 50–70 daily unstructured hiring signals into schema-validated jobs and ranks them with embeddings; 100+ concurrent users, ≈60% lower feed latency; browser-extension autofill, BYOK gateway across five model providers.
Relay (Go, FastAPI, Bedrock, pgvector, Celery, Polygon)github.com/shikherjha/relay-api
  • AI returns platform: a Go multi-objective engine routes returns across seven disposition channels; retrieve-then-rerank matching (pgvector HNSW, Bedrock); passport hashes anchored on Polygon.
AxonAI (LangGraph, FastAPI, Pinecone, React, AWS)axon-ai-frontend.vercel.app
  • Five-agent adaptive tutoring system (Tutor, Planner, Evaluator, Memory, Wellbeing) with persistent cross-session memory; 99.9% uptime, 40% lower UI latency.

Skills

Languages
Go, Python, TypeScript, SQL, C++, Java, Shell
Backend & orchestration
Temporal, River, Celery, RabbitMQ, Kafka, gRPC, FastAPI, LiveKit, SIP/WebRTC
Platform
Kubernetes (Kopf, kubebuilder), Helm, AWS (ECS, Lambda, EKS, Bedrock), GCP, Docker, OpenTofu, CI/CD
Data & AI
PostgreSQL (RLS, pgvector), Redis, MongoDB, Elasticsearch, ClickHouse, Pinecone, RAG, LangGraph, Pydantic AI, LLM-as-judge
Reliability
Datadog, OpenTelemetry, CloudWatch, Prometheus, Grafana, Sentry; idempotency and reconciliation patterns

Achievements

  • Amazon HackOn 6.0 National Finalist: 4th nationally among 70,000+ registrants; Top 7 at the finale.
  • Competitive programming: LeetCode Knight (2000, top 3%); Codeforces Expert; CodeChef 4★ (1846); TCS CodeVita R2 rank 895/20K; Meta Hacker Cup rank 1531.
  • Quant / research: Global rank 11/3,000+ in QuadEye Market Data Prediction; technical consultant at Snorkel AI (LLM benchmarking); problem curator at E-olymp.

Education

Guru Gobind Singh Indraprastha University, EDC | B.Tech, Computer Science · CGPA 9.01 / 10Aug 2023 – Jun 2027