Aikyam x Vecton

Phase 1: AI Transformation Workshop

A structured 2-month engagement with four sessions alternating between leadership and all employees. Built on proven change management frameworks — designed to build AI culture from the top down and the bottom up.

• Top-down vision, bottom-up innovation — leadership sets the AI vision and removes barriers, employees ground it in daily workflow reality. The alternating format creates a feedback loop where strategy and operations continuously inform each other.
• Session 1 (Leadership) sets direction → Session 2 (Employees) grounds it in reality → Session 3 (Leadership) refines based on feedback → Session 4 (Employees) owns the execution.
• Based on McKinsey's AI Transformation Framework, Kotter's 8-Step Change Model, BCG 10-20-70 Rule, and the ADKAR Individual Change Model.

Setting the strategic foundation. Secures executive commitment and establishes the guiding coalition.

• AI in Capital Markets briefing — competitive landscape analysis of what peer firms are doing with AI
• AI Maturity Assessment — Gartner's 5-level model to establish a baseline across strategy, data, talent, infrastructure, governance
• Guiding Coalition formation — 3-4 key members: CEO/MD, department heads, and 3-4 influential mid-level champions
• Define the AI Vision — concrete vision for what AI means for Aikyam: which workflows, which departments, what outcomes

Demystifies AI, builds literacy, and captures grassroots insight from the people closest to daily workflows.

• AI Literacy & Demystification — hands-on introduction using real firm documents: equity research PDFs, SEBI filings, client communications
• Capability Framework — Green (AI does this today), Yellow (AI assists, human verifies), Red (AI cannot do this yet)
• Use Case Discovery — cross-functional teams identify pain points using design thinking. Each team produces candidate use cases with estimated impact
• Feedback Collection — structured capture of employee concerns, ideas, and enthusiasm levels. Feeds directly into Session 3

Leadership reviews employee feedback, refines priorities, and aligns resources.

• Employee Feedback Review — synthesized presentation of Session 2 findings: common pain points, highest-excitement use cases, resistance patterns
• Use Case Prioritization (RICE) — all use cases scored on Reach x Impact x Confidence / Effort, plotted on Value vs Feasibility matrix
• Resource Alignment & Success Metrics — specific KPIs per initiative, departmental champions, data access requirements
• Responsible AI & Compliance — review against Deloitte's Trustworthy AI framework: fairness, reliability, privacy (DPDP Act), security

The organization receives the prioritized roadmap, AI champions are activated, and governance is established.

• Roadmap Presentation — finalized AI roadmap shared with all employees: which initiatives move forward, timelines, department involvement
• AI Champion Network — 1 champion per department (5-7 across the firm) with defined roles: liaison, feedback collector, first point of contact
• Phase 2 Path & Expectations — which products get built, in what order, and what each department should expect
• AI Governance Framework — data access policies, vendor management, responsible AI principles, escalation paths, review cycles

• Current State Assessment — AI maturity scores, department workflow analysis, data readiness heat map, talent gap analysis
• Use Case Registry — full backlog scored on RICE framework, Value vs Feasibility matrix, compliance assessment
• Impact Assessment — time savings potential per initiative, production feasibility, expected outcomes
• Phase 2 Roadmap — detailed roadmap with milestones, AI operating model design, champion network, governance
• Executive Summary — board-ready brief with impact projections and recommended next steps

Distressed Asset Intelligence & Promoter Profiling

Given a company name or CIN, the system auto-assembles a comprehensive dossier: MCA filings, NCLT case status and timeline, promoter-director network mapping (all companies via DIN database), litigation history from eCourts, credit defaults from RBI willful defaulter lists, and lender exposure. Output is a structured report with red flags auto-highlighted.

• Entity resolution — linking promoter names across MCA, eCourts, and news (different spellings, similar company names)
• Knowledge graph — mapping promoter-director-company relationships, identifying circular holdings and related-party networks
• LLM extraction — summarizing legal proceedings, extracting key dates and outcomes from court orders
• Pattern classification — auditor resignation + covenant breach + promoter pledge increase = high risk

• No funded SaaS player builds this for Indian stressed assets
• Data is messy and fragmented — which IS the moat
• Manual dossier takes a team 2-3 weeks per company
• Every new NCLT case is a new revenue event

FPI Intelligence & Enrichment Platform

Takes their existing 11,725 FPI Excel, enriches each entity with SEBI registration details, custodian info, AUM from NSDL monthly data, investment style (sectoral allocation), management personnel, and contact information. Searchable interface with change-detection alerts.

• Entity resolution — matching inconsistent Excel naming to SEBI/NSDL official records
• LLM extraction — management personnel from SEBI registration documents
• LinkedIn matching — automated profile identification via licensed API
• Change detection — flags meaningful shifts (custodian change, AUM movement)
• Outreach generation — personalized drafts based on FPI investment patterns

Focus on 500-1,000 relevant FPIs (400M-1B AUM range) rather than all 11,725. Go deeper rather than broader. Change detection is the real product, not the initial enrichment.

Resume Shortlisting & Talent Matching

AI-powered recruitment platform that automates resume shortlisting and talent-to-JD matching. Ingests resumes in bulk, parses qualifications, skills, and experience, then scores and ranks candidates against specific job description requirements. Designed for capital markets hiring where domain-specific terminology and regulatory knowledge (SEBI, NISM certifications, specific product expertise) are critical differentiators.

• NLP-based resume parsing — entity extraction of qualifications, certifications, skills, employment history from diverse formats (PDF, Word, LinkedIn)
• Semantic JD matching — contextual fit beyond keyword matching. "NCLT resolution experience" matches "stressed asset management expertise"
• Automated shortlisting — configurable thresholds for must-have vs nice-to-have. Ranked candidate list with match scores and rationale
• Skill gap analysis — identifies what each candidate brings vs what the role requires. Highlights training needs and growth potential
• Bias reduction — structured evaluation criteria ensure consistency. Removes unconscious bias by focusing on skill-to-requirement alignment

• Reduces manual screening time by 70-80% per hiring cycle
• Semantic understanding catches candidates that keyword filters miss
• Scales hiring capacity without proportional headcount increase in HR
• Brings consistency and objectivity to initial screening across all positions

MF Portfolio Construction & Rebalancing Tool

Tool for wealth RMs: given a client's risk profile and existing portfolio, recommends target allocation, identifies fund overlap (many clients hold 15+ funds with 60% overlap), generates quarterly rebalancing suggestions with RM-presentable reasoning.

• Portfolio overlap detection using fund-level holding data
• LLM generates client-facing rebalancing narratives
• Risk questionnaire to SEBI-compliant category mapping

Treasury Dashboard with ALM Monitoring

Ingests Aikyam's loan/facility data and investment book. Computes blended cost of borrowing vs GSEC benchmarks, bank-wise exposure concentration, asset-liability maturity mismatches, covenant compliance. Alerts on threshold breaches.

• LLM extracts key terms from sanction letters and facility agreements (interest rate, tenure, covenants, reset dates) — documents are unstructured and vary by bank
• LLM generates monthly treasury summary narratives
• Anomaly detection on blended cost movements

Note: Narrow audience (2-3 people in treasury). Quick to deliver. Good trust-builder if Treasury head is a key stakeholder.

Post-Trade Risk Alert System

A near-real-time system that ingests trade files (CSV/FIX) from their OMS every 15-30 minutes, applies rule-based checks against 13 monitoring points, and generates prioritized alerts with contextual LLM-written explanations.

• Rule engine — quantitative checks on margin utilization, order-to-trade ratios, position concentration
• LLM narratives — contextual alert explanations with market context
• Natural language queries — "Show me all margin breaches for Desk B in March"
• Anomaly detection — isolation forests identify unusual patterns beyond rule-based checks

1. Free Cash vs Total Collateral
2. Integrated Surveillance Dashboard
3. Client Financials vs Trading Activity
4. Global Fat Finger / Algo Error Analysis
5. Trading Activity vs Limits Applied
6. DMA to Non-DMA Ratio
7. Order to Trade Ratio (Algo Trades)
8. Bandwidth Utilisation (Exchange-wise)
9. Order Rejection Analysis
10. Market Movement vs Peak Margin
11. Scenario Analysis by Position Type
12. Client Position vs Physical Delivery
13. Expiry vs Non-Expiry Margin

Equity Research Analysis

End-to-end equity research document and presentation generation platform. Takes company filings, quarterly results PDFs, and financial data as inputs, then generates structured research documents and presentation-ready slides. Analysts upload a company's quarterly filing and receive a comprehensive draft research note in minutes — company overview, key financials, ratio analysis, peer comparison framework, and areas requiring deeper investigation.

• RAG pipeline + LLM extraction — parses quarterly filings, annual reports, and investor presentations. Extracts financial data, management commentary, and key disclosures
• Research note generation — structured draft notes: company overview, financial analysis (revenue, margins, cash flows), ratio analysis (PE, EV/EBITDA, RoE), peer comparison
• Presentation slide creation — presentation-ready slides with key charts, data tables, and narrative summaries formatted to institutional standards
• Consistency & quality standards — ensures all output follows Aikyam's formatting guidelines, terminology, and quality benchmarks

• Reduces routine research drafting from hours to minutes per company
• Maintains institutional quality standards and formatting consistency across all output
• Allows senior analysts to focus on judgment, novel analysis, and client relationships
• Slide generation eliminates the separate step of converting research into presentations