LanceSoft, Inc.
Website:
lancesoft.com
Job details:
Title: Technical Lead / Solution Architect — TGM Market Intelligence Agent
Location : Offshore India- Bangalore
Duration : 3 Month
Role Purpose
- A three-hat role by design: Solution Architect — own the HLD/LLD derived from Architecture v1.0, close the open technical decisions (database/vector engine, real-time access pattern, connector approvals), and keep the architecture coherent as it meets reality. Lead Developer — personally build the hardest components: the agent pipeline, the claim store and diff engines, the graph traversal with evidence gating, the retrieval layer, the compliance gate. Technical Lead — set the engineering bar for the squad, review everything that ships, and represent technology in governance forums. This is a player-coach role for someone who still wants to write the most important code themselves while owning the technical outcome end to end.
Required Expertise
- 10+ years of software engineering with genuine production ownership, including 3–4 recent years designing and shipping LLM-based systems (RAG/retrieval, agentic workflows, structured outputs, evaluation) or comparable large-scale data platforms — and the judgment to know when not to use a model.
- Solution architecture in a regulated enterprise: security review, data classification, audit and model-governance processes; banking experience strongly preferred.
- Deep data engineering foundations: Azure data stack (Databricks/Spark, ADF), PostgreSQL at depth (schema design, indexing, pgvector), API and integration architecture; Oracle exposure useful given the open engine decision.
- Grounded-generation discipline: citations, evidence gating, hallucination control and evaluation design — this platform's credibility rests on every sentence being traceable to a source.
- Understanding of financial markets — rates, FX, commodities, credit; how central-bank communication and data releases move markets; how a trading floor consumes information. Prior work in or delivery into a treasury/markets or trading-office environment is a strong plus.
Required Skills
- Expert Python; production LangGraph/LangChain or equivalent agent-orchestration experience; prompt architecture and structured-output design against enterprise model services.
- Vector and hybrid retrieval implementation; embedding pipelines; reranking and retrieval evaluation.
- Advanced SQL and data modelling, including temporal/versioned data and graph-in-relational patterns (recursive traversal); vector index operations.
- Competent TypeScript/React — able to review and unblock front-end work (shadcn, TanStack Query, Zustand) even if not the primary builder.
- CI/CD, observability and operational design for a system with a hard morning SLA; secure-by-design habits (secrets, least privilege, data classification).
- Strong written architecture communication — HLDs, ADRs and governance-forum material that non-engineers can follow.
Preferred Tools & Certifications
- Microsoft Certified: Azure Solutions Architect Expert (or equivalent cloud architecture credential); Databricks certification.
- Familiarity with Oracle AI Database (26ai) property graph and vector features; Redis; IBM API Connect; MCP-style tool boundaries for AI systems.
- TOGAF or comparable architecture framework exposure (lightweight application expected).
- Financial-markets credentials (CFA Level I, FRM Part I, CISI) welcomed, not required.
Education
- Bachelor's or Master's degree in computer science, software engineering, or a quantitative discipline.
Click on Apply to know more.
